Synapse - Africa’s 4IR Trade & Innovation Magazine - 2nd Quarter 2021 Issue 12

2nd QUARTER 2021 | ISSUE 12

SYNAPSE Africa’s 4IR Trade & Innovation Magazine

NVIDIA INCEPTION: 3 African Startups Accepted into the Programme

TUNBERT: 1st AI-based

Tunisian Dialect System

3 AFRICAN STARTUPS using AI, Data Science for Financial Inclusion

AFDB BACKS AI-BASED

National Consumer Management Systems

HOW THE PANDEMIC Gave Birth to SA’s latest 4IR SaaS platform

LACUNA FUNDS DATASETS for Low Resource African Languages

Zindi is building AI in Africa

26,000+ data scientists 26% female

35+

21%

25 - 34

51%

18 -24

28% 76% users in 45 African countries

$175,000+ awarded in 60 competitions 40 awesome ambassadors in 20 countries 140,000+ pageviews per week 5,000+ Twitter followers

Contents SYNAPSE | ISSUE 12 | 2nd QUARTER 2021

p27 RPA: The Next Chapter In the Automation Story

p29 UNESCO launches AI Needs Assessment Survey in Africa

p53 Optimise & Streamline Your Workspace Management

4 AfDB $1m Grant for AI-based National Consumer Management Systems 5 Hyperautomation: A Case Study 6 All You Need To Know About The EU’s DIGILOGIC initiative 8 Human In The System: Understanding Customer Behaviours 10 These 3 African Startups Are Using AI, Data Science to Disrupt Fintech 11 Why Kenya’s Ajua acquired AI/ ML fintech startup WayaWaya 12 Lacuna Fund invests in Datasets for Low Resource African Languages 14 TunBERT: 1st AI-based Tunisian Dialect System 15 SA Team Places 2nd at 2021 Imagine Cup Junior Virtual AI Hackathon, Girls Edition 16 Meet UCT’s 1st Google Research Scholar Program Recipients 18 How The Pandemic Gave Birth to SA’s latest 4IR SaaS platform 20 NVIDIA unveils its 1st Data Centre CPU 21 Putting AI Into The Engine Room 22 Willis Re Launches new SA Hail Catastrophe Risk Model 23 5 Steps to building a People Analytics Function From The Ground Up 24 How CompariSure’s Conversational AI is Driving Digitisation Within the Insurance Industry 26 Liquid Telecom’s Rebrand 27 RPA: The Next Chapter In the Automation Story 28 UmojaHack Africa 2021 29 UNESCO launches AI Needs Assessment Survey in Africa 30 SA’s COVID-19 Third Wavedetecting Algorithm Released

31 NVIDIA Inception: Meet the 3 African Startups Accepted Into the Programme 32 SANRAL explores Machine Learning Applications for Road Safety, Congestion 33 Strathmore Study Lays Bare Gender Inequality in African AI Industry 34 How Quantum Computing Could Propel Us Light Years Into The Future 36 Siemens, CSIR partner to boost SA 4IR skills 37 Wits Announces Team to Advance AI Research in Africa 38 Clevva joins Blue Prism’s Digital Exchange 39 4 Reasons Why You Should Care About AI Governance Now 40 Machine Learning Sandcastles 44 Automation 360: Automation Anywhere’s Cloud-Native Platform for Intelligent Automation 46 How the AU, Africa CDC will take On COVID-19 Through AI, Big Data 47 DeepMind Establishes Scholarship for Wits Masters Students 48 Why 20k+ Developers from Emerging Markets Signed Up for GTC 49 Innovation Factory (Africa) Challenge 50 How IBM Wants to Accelerate DX With Latest Breakthroughs in Hybrid Cloud, AI Capabilities 52 Invisio AI wins at the 2020 SAB Foundation Social Innovation & Disability Empowerment Awards 53 Optimise & Streamline Your Workspace Management

2ND QUARTER 2021 | SYNAPSE

1

Editor's Notes Daniel Mpala, Editor and Head of Show Production

C

onversational AI solutions are proving a crucial tool to have in your toolbox if your firm cares about customer experience. The interest and investment in this space is a sign of just how important these have become in the current pandemic, and will likely be in a post-COVID-19 world. More importantly, these tools and the amount of research into translation models catering to low-resourced African languages is giving new promise to how these technologies can help include more people in the digital economy. Recently Africa scored a big win with grassroots Natural Language Processing (NLP) research community Masakhane bagging the Wikimedia Foundation Research Award for its work, which has led to the launch of its Machine translation service which can translate Yoruba, Lingala, Tshiluba, Igbo, Shona and Setswana. Synapse Issue 12 will focus on some of the major investments, startups and news around Conversational AI and NLP innovations in Africa over the last quarter. With less than 15 weeks to go to AI Expo Africa 2021 ONLINE, have you registered for your free pass yet? This year we have several new local and international exhibitors showcasing avant-garde 4IR tech and a great lineup of thought leaders across the AI, Data Science and 4IR smart tech scene that you’ll definitely wan’t to watch speak.

Speaking of thought leaders, don’t miss out on some free insights in this issue. We hope you enjoy this edition.”

JOIN US - ONLINE - ANYWHERE

www.aiexpoafrica.com 2 SYNAPSE | 2ND QUARTER 2021

Dr Nick Bradshaw, CEO AI Media Group & AI Expo Africa founder

I

7-9SEP2021

t’s hammer time at AI Media HQ as we move into top gear on the journey towards AI Expo Africa 2021 - the 4th edition of the show. It’s already seen a lot of early interest for what is shaping up to be our biggest event to date and with a new look, 3-day live online format, kicking off on 7-9th September. Registration interest has already exceeded 5000+ delegates and over 1000+ organisations by end of April which is a new milestone for us. It’s also been great to many new exhibitors and vendors joining the show. April and May saw us launch The Deepfake Africa Challenge on the Zindi platform to draw attention to this medium and create debate about its use. We also attended the Nvidia GTC Event which saw a massive surge in emerging market delegates and a sure sign that interest in AI is growing in our region. We also launched the ITU Innovation Factory Challenge (Africa), in partnership with AI Centre of Excellence (AICE) in Kenya, focusing on AI4Good innovation and entrepreneurism. We have a load of new announcements coming in the next three months in the run up to the show which is a really busy time for the team – some midnight oil will be burned thats for sure! Daniel Mpala, Editor of Synapse, has yet again delivered a great 12th edition of the magazine which is now almost 3 years old. Time has flown since we launched the magazine in 2018 and have covered over 100 stories from contributors across the Africa region.

We want to thank all our supporters and contributors who have helped get the magazine this far!

INVESTMENT

INNOVATION

AFDB PROVIDES $1M GRANT for AI-

based national customer management systems in Ghana, Rwanda & Zambia

T

he African Development Bank (AfDB) announced in March that its Board of Directors had approved a $1.024-million grant for AI-enabled systems to process customer complaints on behalf of the national banks of Ghana, Rwanda, as well as for the Competition and Consumer Protection Commission of Zambia.

The grant was made from the Africa Digital Financial Inclusion Facility (ADFI), a special fund dedicated to accelerating digital financial inclusion across the continent.

The AfDB said the project will establish a complaints-handling system for the financial regulators, using multilingual chatbots and artificial intelligence that will interface with key financial services providers in the three countries.

The system will incorporate key local languages for ease-of-use, record customer complaints, including audio complaints from those unable to read and write, and track their resolution.

The project is expected to improve the tracking of customer complaints made to financial services providers, strengthen the support for marginalised groups -- which will build confidence in the use of financial services -- and improve the collection

4

SYNAPSE | 2ND QUARTER 2021

of consistent data to be used for the development or improvement of consumerprotection policies.

ADFI co-ordinator Sheila Okiro pointed out that facilitation of sound policies and regulations, including those that enhance consumer protection and catalyse financial inclusion, is a key man-date for ADFI. “With the proliferation of digital financial services, the financial industry needs innovative mechanisms for customer recourse and tracking for regulators. The Sinitic project is one such solution,” added Okiro.

The system will be developed by Sinitic Africa in collaboration with BFA, a consultancy firm specialising in humancentred design and DFS regulation. Sinitic Africa is a subsidiary of Sinitic Inc., a financial technology firm based in Canada.

The two companies have already worked together to develop and successfully deploy a similar project for the Philippines’ central bank.

The Sinitic solution will be deployed in the three target countries in the following languages: Kinyarwanda, Swahili, French and English in Rwanda; English and Nyanja/ Chewa in Zambia; and English and Twi in Ghana.

Grassroots NLP community Masakhane wins Wikimedia Foundation Research of the Year award Masakhane, a grassroots organisation which aims to strengthen and encourage Natural Language Processing (NLP) research in African languages for Africans by Africans was in April awarded the Wikimedia Research Award of the Year for its paper Participatory Research for Lowresourced Machine Translation: A Case Study in African Languages.

Masakhane received the award together with Kay Zhu, Dylan Walker and Lev Muchnik for their paper Content Growth and Attention Contagion in Information Networks: Addressing Information Poverty on Wikipedia. The award was presented virtually to Masakhane by Wikipedia cofounder Jimmy Walles.

Machine Translation (MT) is crucial for information accessibility and communication worldwide. Despite this, MT is still centred around a few highresources languages, widely excluding African languages.

ADVERTORIAL

HYPERAUTOMATION:

A case study

R

ecently, we had an opportunity to help one of the prominent banks in Indonesia to solve a problem they faced in the cheque clearance process.

The process involves verification of the authenticity of the signature in the cheque, which is a time-consuming process and often overlooked by the clerks. Simple mistakes in signature verification could result in transaction delays and frictions in the bank operations and its customer experience. Our client wanted us to build a solution for solving this challenge. Our answer was Hyperautomation.

As per Gartner’s top 10 strategic technology trends for 2020, Hyperautomation is one of the most trending technologies. In a nutshell, it is the combination of Robotic Process Automation technology with Artificial Intelligence Development. This combination of technologies offers a myriad of benefits, as it enables machines to have the intelligence to automate more complicated tasks.

Robotic Process Automation technology can enable machines to automate rule-based tasks. But, to automate more complex tasks that demand cognitive reasoning, rulebased RPA is not sufficient. For example, an RPA solution can extract the signature and other details from a cheque. But, to verify the authenticity of a signature, the solution needs to have some level of intelligence to compare the sample signature with the verified signature available in the database and make a decision whether the sample signature is authentic or not. And this is where AI comes into play. AI technologies such as Computer Vision, Natural Language

/ By Accubits Technologies /

Processing, Machine Learning, Deep Learning, etc can be employed along with RPA software to automate complex tasks at scale. It can be considered as a level above RPA, further substituting the involvement of humans in both physical and digital tasks.

The benefits offered by hyper-automation are huge. It has been steadily gaining popularity as companies have started prioritising increased ROI above all else.

According to Coherent Market Insights, the international hyperautomation market is expected to develop at a CAGR of 18.9 % during 2020-27 with widespread digitalisation of outdated manufacturing plants being the key contributor to the numbers.

Hyperautomation technology helps companies to amplify their digital journey.

It also lays down a strong foundation for the production of crucial innovations and enhancements in business growth. The technology mostly results in the making of a digital twin of an enterprise/organisation, letting them envisage its functions, procedures, and key performance indicators to drive more value. One of the most sought after use cases of hyperautomation is that It can connect various departments in an organisation. Such tools can help in connecting data, processes, and operations from various areas. The combination of RPA, intelligent tools, and software robots with users across all organisational core processes makes it easier for users to collaborate.

Considering the challenges faced by our client, we developed a computer visionbased signature forgery detection system.

Using computer vision, the solution was able to learn textual patterns in scanned cheques.

The system uses deep learning algorithms to compare it with the original signature to identify even the minutest variations.

However, there are situations when one might not be able to put a signature the exact same way as the sample available with the bank. We designed the AI model to identify and measure this mismatch percentage and determine whether it is, in fact, a genuine signature that simply has a minor difference from the original one.

This greatly improves the reliability of such systems as it does not cause inconvenience for genuine signatories. The solution was able to reduce the manpower requirement in one branch from 12 employees to 2 employees and reduced the cheque signature verification time by 70%.

Hyperautomation eradicates bottlenecks across the chain of operations, optimises processes, and reduces the number of timeconsuming manual tasks. This ensures that your employees are more productive, driven, and motivated to take on more work.

2ND QUARTER 2021 | SYNAPSE

5

NEWS

DIGILOGIC, a smart logistics

focused pan European & African Digital Innovation Hub launched

The European Commission has granted €1.9-million to DIGILOGIC, a three-year European Union Horizon 2020 project which aims to boost cooperation and long-term, sustainable partnerships between European and African Digital Innovation Hubs (DIHs), paving the way for innovators, startups, and SMEs to jointly develop smart logistics solutions in close cooperation with industries and investors.

T

he project, which was launched in January, is deployed by a consortium comprising three DIHs in Europe - Digital Hub Logistics Dortmund (Germany), VTT (Finland) and Friuli Innovazione (Italy); two DIHs in Africa - MEST (Ghana) and BongoHive (Zambia) , as well as system change facilitator Endeva (Germany) and an SME Prototipi (Nigeria).

The consortium said in a joint statement in February that the seven partners will together foster the adoption of emerging

6

SYNAPSE | 2ND QUARTER 2021

technologies such as Cloud Computing, Big Data, Augmented and Virtual Reality, Machine Learning, Blockchain, Artificial Intelligence (AI),Smart Devices, Internet of Things (IoT) and Intelligent Transport Systems (ITS) for smart logistics solutions, through the deployment of a dynamic and impactful knowledge transfer process and a targeted implementation programme across Europe and Africa.

DIGILOGIC will work with digital innovation hubs to foster the adoption of emerging technologies for innovative solutions, products and services. These

organisations, it said, create the conditions for -- and subsequently nurture, support and promote -- innovation, especially by SMEs, young innovators, startups and mid-caps. The digital innovation hubs also provide access to technology-testing, skills development, financing advice, knowledge sharing, collaborative projects, brokerage, market intelligence and networking opportunities.

The European Commission is facilitating the creation of networks of DIHs which can share nodes/technologies platforms, as well as hard and soft competencies

NEWS

DIGILOGIC go-to-market focus eLearning

to cope effectively with the EU’s digital transformation across all aspects of the economy. In Africa, there is an increased commitment to coordinate efforts across ecosystems that has led to the establishment of several regional and continental networks of hubs: a recent report

by Briter Bridges and the GSMA Ecosystem Accelerator Programme maps over 600 tech hubs on the continent, joining forces in around 15 networks.

DIGILOGIC will act as an ecosystem of ecosystems, mapping the existing players, leveraging on DIHs’ competencies and network and analysing the levers of change which should be addressed to exploit innovation, facilitate collaboration and ultimately create market and uptake opportuni-ties for innovators, SMEs and startups in smart logistics from both continents.

DIGILOGIC Smart Logistics Technology Radar DIGILOGIC partners see the horizontally connecting logistics industry at the converging point of interest and priorities

“DIGILOGIC will work with

digital innovation hubs to foster the adoption of emerging technologies for innovative solutions, products and services

”

for digital innovation for social and business development; a crucial node for Europe’s and Africa’s sustainable prosperity. The logistics sector is facing a plethora of challenges, both from competitors as well as disruptors.

It is therefore a sector that cannot afford to not keep innovating, to secure a free and independent logistics industry.

The innovation uptake by applying ICT and other digital technologies to logistics will have a knock-on effect of benefiting other industries (from agriculture to manufacturing, from local trade to ecommerce, from health care to mining) and will provide substantial support to both macro- (e.g. large scale freight movements by air, sea and land, added-value services like warehousing, assembling among others) and micro-/ small- businesses growth in areas such as “last-mile” deliveries.

Moreover, the ongoing COVID-19 pandemic is showing the essential role played by logistics in sustaining the functioning of our societies and delivering vaccines, and emergency supplies, while revealing the fragile nodes of global supply chains. Technology is helping our societies to remain connected at the time of social distancing and e-commerce stepped in significantly in our daily lives.

The consortium pointed out that now, more than ever, data, knowledge and competencies sharing are the combustible for innovation and recovery: “There is no time to lose and nobody can afford to be isolated”.

DIGILOGIC is working on a smart logistics technology radar, including the latest technical developments and industry needs, with a deep dive into specific European and African challenges and available solutions, to guide the mentoring and go-to-market learning programme.

In June 2021, DIGILOGIC plans to launch an interactive eLearning platform with on-demand and live business, design thinking and smart logistics ICT modules. The eLearning platform will be free of charge and accessible at any time with on-demand and live webinars, documentation, dedicated thematic course, exercises and assignments to enable learners to follow their learning path as per their needs, motivation and time availability. Through the same platform, smart logistics experts will offer group and individual mentoring to innovators. The DIGILOGIC eLearning platform will also represent the place for peer-to-peer learning, where innovators can network and find the right partners to develop their proposals to DIGILOGIC’s so called Challenges.

DIGILOGIC Challenges for European and African innovators In May 2022, DIGILOGIC plans to launch opportunities for innovators in Europe and Africa to solve four challenges related to the improvement of logistics on both continents. DIGILOGIC will define the topics of the four challenges, based on the real needs and pain points of SMEs, industries, and local governments. There will be opportunities to solve challenges ranging from logistics needs arising from ports to distribution centres, from transport organisations or local governments, solving bottlenecks at cross-border processes or last-mile delivery, at the warehousing stage or distribution level.

Applicants will be invited to propose effective emerging technologies solutions (such as IoT, AR/VR, AI, blockchain, etc.), effective deployment, demonstrate a combination of ICT and business skills and show the successful cooperation of innovators from Europe and Africa. The selected proposals (up to 12) will be offered a one-year mentoring and coaching programme which will kick off with a three-day Bootcamp in Europe and continue with ongoing virtual and physical mentoring and access to technology infrastructures made available by each DIH. The best projects will be part of DIGILOGIC’s final Demo Day, co-located with a major innovation event in Europe or Africa.

2ND QUARTER 2021 | SYNAPSE

7

ADVERTORIAL

HUMAN IN THE SYSTEM:

Understanding customer behaviours with ecosystem.Ai

T

here’s a young man in a crowd of people in a small village. He’s holding a saxophone and he’s dying to play it. The problem this young man has is that he’s missing a vital part of the instrument - the mouthpiece. He’s been searching far and wide, to no avail, until one day he comes across an old woman who offers him a reading of his future. She informs him that she foresees that he will find what he is looking for soon enough, he glances down at the sax and fiddles with the hole where the mouthpiece should be. This old woman deduces what his heart most desires. She calls upon a merchant she knows well, and has worked with before, and tells him that if he can find the instrument’s missing part, they may just have a new customer.

In the time when humans consulted oracles, and merchants knew their customers by name, trade was conducted almost entirely by means of personalisation and individual communication. These early forms of business conduct relied heavily on knowing your customer, and the same applied to matters of prediction.

While merchants offered handpicked wares for their best customers, misty women

8

SYNAPSE | 2ND QUARTER 2021

in large shawls offered grand futures to those they knew best. When the two worked together, they had the perfect business.

Both trade and prediction, however, needed to evolve to accommodate customer base growth; and thus the close relationship between the offerors and the offerees was lost.

The loss of individualisation though, did not change the fact that customers are unique - despite having been treated as such in broad spectrum sales. Business history has now come full circle. After years of data collection, it is now time to return to a place of personalisation. ecosystem.Ai, with the capabilities of machine learning, has found a way to extract human behaviour from data. Allowing companies to form stronger relationships with their customers, by understanding the clues in their behaviour. By being able to know your customer is behaving like a Jazz King, and communicating in a way that will resonate with them. ecosystem.Ai has been working tirelessly to create an effective set of tools, which companies can use to both understand, and engage, with their customers on a deeper level. Using the theory and practices of Computational Social Science, in conjunction with contemporary psychological constructs,

the chances of seeing the human in your data is not just a business dream. The practical implementation of dynamic behavioural analytics and segmentation, helps to progress common business practices beyond the static boundaries of mathematical analysis and categorisation. ecosystem.Ai offers a tripart selection of products: The ecosystem.Ai Workbench is a versatile, customisable user interface that reduces the complexity of machine learning processes. The Workbench is bolted directly into a business’s internal structure, maintaining exclusive privacy on company/ customer data. There are a vast number of capabilities available in the ecosystem.Ai workbench, including: project creation and management, data science engineering, model creation and deployment; amongst others.

The second of the products is a series of pre-written algorithm Modules. These Modules can be used to solve common business problems (such as churn interventions or offer recommenders) using ecosystem.Ai’s behavioural constructs designed for a sector or exclusively for a company. Check out ecosystem.ai/#modules to find out more.

The third product is The Client Pulse Responder which is a fully configurable and automated application that uses continuous re-enforcement learning. The Client Pulse Responder is the executable environment that ensures models are deployed, and results are seen. It is the internal core of ecosystem.Ai’s products, providing the space in which business problems are transformed into actionable solutions.

The set of tools that ecosystem.Ai has developed equip clients with the knowledge and capacity to learn more about their customers. All of these elements help businesses see that collections of events in an individual’s life are unique enough to provide a view of the human in the data. ecosystem.Ai are the creator of the tools that contemporary merchants and oracles use in their merged business. By using the combination of business practice, the human sciences of social theory and psychology, and computational power; it is possible to reform the close relationship between company and customer. Understanding customer dynamics through data, offers the unique insight needed to know that your customer is behaving like a Jazz King.

NEWS

CATAPULT: INCLUSION AFRICA 2021:

Meet the 3 African startups using AI, Data Science to disrupt fintech Three African startups using artificial intelligence (AI) and Data Science to bring about financial inclusion on the continent — Nokwary, Mosabi and Mipango— were among 14 companies selected to participate in the Luxembourg House of Financial Technology (LHoFT) CATAPULT: Inclusion Africa 2021 programme.

10

SYNAPSE | 2ND QUARTER 2021

T

he fintech startup development initiative — which was launched in 2018 and then subsequently held in 2020 — was designed for companies working on financial inclusion in Africa. In addition, the programme also aims to build bridges between Africa and Europe.

This year’s programme, which was held online between 19 and 28 May, provided the 14 participants with intensive mentoring, coaching, peer to peer learning and dedicated workshops. This year’s programme also features sessions on business model mapping, investment readiness, funding and capital raising, social impact, scaling strategy, building teams, operational management, advisory meetings with investors, and pitch development.

Here’s how the three are using AI and Data Science to bring about financial inclusion on the continent: Nokwary (Accra, Ghana): Nokwary uses AI to break barriers to digital access, connect people and bring convenience to consumers. It’s expertise in speech recognition and conversational user

interfaces has enabled the startup to develop systems that enable people who can’t see as well as those that can’t read and write to interact with banking apps and systems through spoken commands.

Mosabi (Freeetown, Sierra Leone): Mosabi links fintech and edtech with embedded, gamified up-skilling for Africa’s financial products and platforms. This helps underserved citizens improve decisions and behaviours on their businesses and money. Mosabi harnesses data insights from user profiles and journeys, to match them to best fits across a marketplace of digital financial services.

Mipango (Dar es Salaam, Tanzania): Mipango is a personal finance mobile application that enables users to manage their income, expenses, savings, financial targets, help and offerings.

Mipango’s app uses AI to guide users to better financial management and financial deals. In addition, the app also helps users with personalised financial advice and investment opportunities.

INDUSTRY

WayaWaya founder Teddy Ogallo and Ajua CEO Kenfield Griffith

Kenya’s Ajua acquires AI/ ML fintech startup

WAYAWAYA

enables businesses access to digital payments using a unique USSD code, CRM tools, customer feedback channels, debt management and tracking, business and product promotions through mobile and social media channels. Through its new product roll-out with MTN, Ajua is generating more data for its thousands of users, much of which can now be better automated and monetised through the products and services WayaWaya has built, including cross-border digital transfers, payments services and intelligent finance bots.

Griffith said the acquisition of WayaWaya is an important milestone for Ajua, as the firm makes a significant leap in ensuring the customer experience journey for businesses across the continent is seamless.

“Integrating WayaWaya’s technology significantly complements our product suite and gives us the ability to automate our clients’ businesses and grow their revenues, which is an extremely powerful proposition for our customers of all sizes, across Africa. From our experience in this area, we understand the CX fundamentals that drive growth for our customers and we want to bring this intelligence to SMEs across the continent,” he added.

“The additional reach this acquisition brings allows Ajua to scale significantly within the SME vertical, as we provide our customers today, and in the future, the tools they need to grow in Africa and beyond.

Kenyan firm Ajua — an integrated consumer experience management platform for African business — in April acquired Nairobi-based Artificial Intelligence (AI) and Machine Learning (ML) fintech startup WayaWaya for an undisclosed amount.

W

ayaWaya has developed the Janja platform which enables borderless banking and payments across apps and social media platforms. The platform helps both individuals and businesses with intelligent messaging, across a number of social platforms, including Whatsapp, Facebook messenger, Telegram, and others, and allows its users to automate customer support and make cross-border payments.

The acquisition will see WayaWaya founder Teddy Ogallo join Ajua as VP of Product APIs and Integrations. Ajua said in a statement that the deal will enable the consumer experience firm to automate much of the customer experience journey by integrating the janja.me product into its product stack, closing the customer experience loop as

the smart AI and ML built by WayaWaya gives SMEs the ability to automate responses and give the customer what they want, when they want it.

Launched in 2012 as mSurvey by founder and CEO Kenfield Griffith, Ajua aims to solve the customer experience gap for businesses on the continent to drive business growth. Ajua combines technology with customer experience, and has built a number of products that deliver real-time customer feedback at the point of service, for small and large businesses across Africa, powering customer experiences for businesses with over 45-million customers.

Current Ajua partners and clients include GoodLife Pharmacy, Standard Chartered, FBNQuest, Safaricom, Total, Coca-Cola and Java House.

Ajua’s acquisition of WayaWaya comes hot on the heels of Ajua’s partnership with Nigeria’s MTN for MTN EnGauge, an agile application that offers innovative customer engagement solutions. Ajua explained that the platform

We continue to be bullish on the point that customer experience and customer engagement are the engine for growth for businesses across the continent and they are disciplines that are critical factors in driving productivity and revenue growth,” said Griffith.

Ogallo said Ajua’s focus on introducing and scaling customer service and customer experience for the continent, as well as how they help businesses deliver excellence for their customers is something he along with his team have long admired. “Seeing how WayaWaya’s technology can complement Ajua’s innovative products and services, and help scale and monetise businesses, is an exciting opportunity for us, and we are happy that our teams will be collaborating to build something unique for the continent,” he added.

2ND QUARTER 2021 | SYNAPSE

11

INVESTMENT

LACUNA FUND INVESTS $1M

in Datasets for Low Resource Languages in Sub-Saharan Africa Lacuna Fund — the world’s first collaborative effort to provide data scientists, researchers and social enterprises in low- and middle-income contexts globally with the resources they need to produce training datasets that address urgent problems in their communities — has invested $1-million in six projects which are creating openly accessible text and speech datasets that will fuel natural language processing (NLP) technologies in 29 languages across Africa.

T

he fund pointed out in a statement in late April that the supported projects will produce text and speech datasets for NLP technologies that will have significant downstream impacts on education, financial inclusion, healthcare, agriculture, communication, and disaster response in Sub-Saharan Africa.

Lacuna Fund explained that the funding recipients will produce training datasets in Eastern, Western, and Southern Africa that will support a range of needs for low resource languages, including machine translation, speech recognition, named entity recognition and part of speech tagging, sentiment analysis, and multimodal datasets. All datasets produced will be locally developed and owned, and will be openly accessible to the international data community.

“With over 50 impressive applications from, or in partnership with, organisations

12

SYNAPSE | 2ND QUARTER 2021

across Africa, there are many more initiatives poised for impact. This movement towards locally developed and owned datasets has only just begun, and with the right support and funding these initiatives will unlock the power of AI to deliver new social sector solutions and increase the presence of African countries on the international data map,” Lacuna Fund stated.

Also commenting in the same statement, ABSA Chair of Data Science at University of Pretoria Vukosi Marivate drew attention to how the South African government has been using chatbots to provide daily updates on COVID. “Right now, translating those updates to Latin languages is really easy, but the datasets necessary to translate those updates to a range of African languages don’t exist, which means that the government isn’t currently able to communicate with many of its people in their native languages. That is one of the many examples of why we need this work now,” explained Marivate.

Meet the recipients Building an Annotated Spoken Corpus for Igbo NLP Tasks: This project addresses the gap in the availability of an Igbo spoken corpus for NLP tasks. Existing corpora—such as the Igbo web Corpus (IgWaC) and literary, religious and grammar texts—are either unannotated or not archived for research and NLP tasks. This study will create an annotated 1000-sentence corpus and 25 hours of unannotated audio data to launch an open access spoken corpus that would be available for research and NLP tasks.

Data will be gathered from oral narratives and live Igbo news. Ethnographic interviews will be used to collect data that covers several domains of the Igbo life such as marriage, religion, language, burial, education, security, and trade. To ensure adequate representation, balance, and homogeneity, data collection will take place in the five south-eastern states where Igbo is predominantly spoken, and the team will recruit 50 different language speakers across the states to provide audio data. Igbo news recordings will be acquired from the Federal Radio Corporation of Nigeria across the five states.

Igbo NLP Tasks Project Team member Gerald Nweya from the University of Ibadan said the team is excited to embark on this project due to the impact it will have on the NLP community as it it particularly concerns the Igbo language. “The need to build an annotated corpus of contemporary Igbo is one that is long overdue. It could be very interesting to study the language

INVESTMENT

from naturally occurring contexts such as narratives, stories and conversations.

Therefore, we are both overjoyed and grateful to the Meridian Institute for giving us this unique opportunity through Lacuna Fund. We are very hopeful that this will serve as a springboard for the use of Igbo for NLP tasks and other applied linguistic research,” added Nweya.

Masakhane MT: Decolonising Scientific Writing for Africa: The ability for science to be discussed in local indigenous languages can not only help expand knowledge to those who do not speak English or French as a first language, but also can integrate the facts and methods of science into cultures that have been denied it in the past. Thus, the team will build a multilingual parallel corpus of African research, by translating African preprint research papers released on AfricArxiv into six diverse African languages.

Masakhane MT: Decolonizing Scientific Writing for Africa Project Team member Jade Abbot pointed out that when it comes to communication and education, language matters. “The ability of science to be discussed in local indigenous languages not only can reach more people, but can open up African methodologies and research to the world. We’re exceptionally excited to bring African science to the global community and continue the journey of decolonisation of scientific discourse,” added Abbot.

Multimodal Datasets for Bemba: This project will create the first multimodal dataset for Bemba—the most populous language in Zambia, but one that lacks significant resources. The team will collect visually-grounded dialogues between native Bemba speakers, which will be diarised and transcribed in full. A sample of the data will also be translated into English. The dataset

“This project will lead to a

better understanding of the linguistic structures of 20 African languages from four language families (Afro-Asiatic, English Creole, Niger-Congo, and Nilo-Saharan) and regions of Africa. It will also encourage benchmarking of African language datasets in natural language processing (NLP) research. We look forward to how this initiative will spur NLP research in African universities

”

will enable the development of speech recognition and speech and text translation applications, as well as facilitate research in language grounding and multimodal model development.

Multimodal Datasets for Bemba team member Clayton Sikasote explained that this will be the first multimodal speech dataset created for any Zambian language.

“We are excited about this project because the dataset will enable the development of speech recognition and speech to text translation applications, as well as facilitate research in language grounding and multimodal model development,” added Sikasote.

Named Entity Recognition and Parts of Speech Datasets for African Languages: Currently, the majority of existing Named Entity Recognition (NER) datasets for African languages are automatically annotated and noisy, since the text quality for African languages is not verified—only a few African languages have human-annotated NER datasets.

Likewise, the only open-source POS datasets that exist are for a small subset of languages in South Africa, and Yoruba, Naija, Wolof, and Bambara. This project will develop a Part-of-Speech (POS) and (NER) corpus for 20 African languages based on news data. NER is a core NLP task in information extraction, and NER systems are a requirement for numerous products from spell-checkers to localisation of voice and dialogue systems, conversational agents, and information retrieval necessary to identify African names, places, and people.

“This project will lead to a better understanding of the linguistic structures of 20 African languages from four language families (Afro-Asiatic, English Creole, NigerCongo, and Nilo-Saharan) and regions of Africa. It will also encourage benchmarking of African language datasets in natural language processing (NLP) research. We look forward to how this initiative will spur NLP research in African universities,” stated Masakhane.

Building NLP Text and Speech Datasets for Low Resourced Languages in East Africa: The project will deliver open, accessible, and high-quality text and speech datasets for low-resource East African languages from Uganda, Tanzania, and Kenya.

Taking advantage of the advances in NLP and voice technology requires a large corpora of high quality text and speech datasets. This project will aim to provide this data for these languages: Luganda, Runyankore-Rukiga, Acholi, Swahili, and Lumasaaba. The speech data for Luganda and Swahilli will be geared towards training a speech-to-text engine for an SDG relevant use-case and general-purpose ASR models that could be used in tasks such as driving aids for people with disabilities and

development of AI tutors to support early education.

Monolingual and parallel text corpora will be used in several NLP applications that need NLP models, including natural language classification, topic classification, sentiment analysis, spell checking and correction, and machine translation.

Open Source Datasets for Local Ghanaian Languages: A Case for Twi and Ga: This project will develop a new speech dataset that makes it possible for Twi (Asante, Akuapim, Fante dialects) and Ga speakers in Ghana with low English literacy to access digital financial services in their native language.

Access to digital financial services will serve as the immediate use case—however, the bulk of the collected data will be additionally useful for other purposes. The team will build a phonetically balanced speech corpus (with transcriptions and rough English translations) that is focused on the financial domain, and since the speech corpus will be phonetically balanced, it should be useful in acoustic modelling for use cases beyond accessing digital financial services.

English illiteracy and low literacy, the Asheshi University and Nokwary team explained, are barriers that keep many Ghanaians from accessing the full benefits of the digital age and in particular, digital financial services. “Advancement in speech and language technology can break this illiteracy barrier but it is impossible to apply these advances to our native languages without datasets in these languages. The funding from Lacuna Fund will enable us to build a dataset in Twi and Ga that we believe will spur AI innovations that will help bring the full benefits of the digital age to all Ghanaians regardless of socio-economic status,” they added.

Lacuna Fund began as a funder collaborative between The Rockefeller Foundation, Google.org, and Canada’s International Development Research Centre, with support from the German development agency GIZ on behalf of the Federal Ministry for Economic Cooperation and Development (BMZ) on this call for proposals.

Earlier this year Lacuna Fund announced that six African projects building agricultural datasets for AI would receive a total of $1.1-million in its first round funding. (Read more about this cohort in this Synapse Q1 2021 article).

Launched in July 2020 with pooled funding of $4-million, Lacuna Fund supports the creation, expansion, and maintenance of datasets used for training or evaluation of machine learning models, initially in three key sectors: agriculture, health, and languages. The initiative has since evolved into a multi-stakeholder engagement composed of technical experts, thought leaders, local beneficiaries, and end users.

2ND QUARTER 2021 | SYNAPSE

13

NEWS

INSTADEEP,

iCompass partner on TunBERT - First AI-based Tunisian Dialect System InstaDeep, an AI startup founded by Tunisian co-founders Karim Beguir and Zohra Slim, and Tunis-based startup iCompass in March jointly revealed a collaborative Natural Language Processing (NLP) project that will lead to the development of a language model for Tunisian dialect, TunBERT. The project will evaluate TunBERT on several tasks such as sentiment analysis, dialect classification, reading, comprehension, and question answering. The partnership aims to apply the latest advances in AI and Machine Learning (ML) to explore and strengthen research in the fast-emerging Tunisian AI tech ecosystem.

“We’re excited to reveal TunBERT, a joint research project between iCompass and InstaDeep that redefines state-of-the-art for the Tunisian dialect. This work also highlights the positive results that are achieved when leading AI startups collaborate, benefiting the

Tunisian tech ecosystem as a whole,” said InstaDeep CEO Karim Beguir.

Bidirectional Encoder Representations from Transformers (BERT) has become a state-of-the-art model for language understanding. With its success, available models have been trained on Indo-European languages such as English, French, German etc., but similar research for underrepresented languages remains sparse and in its early stage. Along with jointly writing and de-bugging the code, iCompass and InstaDeep’s research engineers have launched multiple successful experiments. iCompass CTO & co-founder Dr Hatem Haddad explained that the collaboration aims to push forward and advance the development of AI research in the emerging and prominent field of NLP and language models. “Our ultimate goal is to empower Tunisian talent and foster an environment where AI innovation can grow, and together our teams are pushing boundaries” said Dr Haddad.

TunBERT is developed based on NVIDIA’s NeMo toolkit which the research team used to adapt and fine-tune the neural network on relevant data to pre-train the language model on a large-scale Tunisian corpus, taking advantage of the BERT model that was optimised by NVIDIA. TunBERT’s pre-training and fine-tuning steps converged faster and in a distributed and optimised way thanks to the use of multiple NVIDIA V100 GPUs. This implementation provided more efficient training using Tensor Core mixed precision capabilities and the NeMo Toolkit. Through this approach, the contextualised text representation models learned an effective embedding of the natural language, making it machine-understandable and achieving tremendous performance results.

Comparing the NVIDIA-optimised BERT model results to the original BERT implementation shows that the NVIDIA-optimised BERT-model performs better on the different downstream tasks, while using the same compute power.

INNOVATION

FIRST FON TO FRENCH

Neural Machine Translation Engine launched edAI researchers Chris Emezue and Bonaventure Dossou have launched FFRTranslate, the first Neural Machine Translation engine from Fon — a very low-resource and tonal language — to French and vice-versa.

Fon shares tonal and analytical similarities with the Niger-Congo languages which include Igbo, Hausa, Yoruba, and Swahili. The engine will promote better communication in Fon and could enable companies to translate texts and messages from Fon to French and vice versa.

Dossou described working on the engine as being “an awesome ride”.

“We’re thankful to everybody that supported us especially our beloved Masakhane NLP family, and the broader

14

SYNAPSE | 2ND QUARTER 2021

NLP community. We believe that this is a huge step toward empowering African endangered languages,” he said in a LinkedIn post announcing the launch.

Other contributors who worked on the project include Fabroni Yoclounon and Ricardo Ahounvlame.Read more about the translation engine in this Synapse Issue 8 article here.

Left to right: Chris Emezue, Bonaventure Dossou

NEWS

SA TEAM PLACES 2ND

at 2021 Imagine Cup Junior Virtual AI Hackathon, Girls Edition Team Cognition, a South African team of Grade 10, Grade 11, and Grade 12 students from different Curro schools around the country won second place at the 2021 Imagine Cup Junior Virtual AI Hackathon, Girls Edition.

T

he challenge, which is known as the “Olympics of Technology” is is a live international event where teams (of 2-5 individuals) compete by trying to solve real-world problems through the use of Artificial Intelligence (AI). The event is sponsored by Microsoft, in partnership with UNESCO.

Imagine Cup Junior 2021, which was held in March, provided students aged 13 to 18 the opportunity to learn about technology and how it can be used to positively change the world. The global challenge is focused on Artificial Intelligence (AI), introducing students to AI and Microsoft’s AI for Good initiatives so they can come up with ideas to solve social, cultural and environmental

“She added that the challenge gave the team a whole new perspective of “new jobs” out there and careers they might look into in the future

”

issues. It’s also a great opportunity to encourage students to develop and practice 21st century skills like communication, collaboration, critical thinking and creativity.

The girls were taken on a practical journey into the heart of AI, to help them develop widely applicable machine learning skills in the context of sustainability, biodiversity loss and climate change. Each team had to present their big idea to a group of ‘investors’ in the first round. A total of 22 teams competed and 11 were chosen to go through to round two of the hackathon.

Team Cognition was pipped by Team Clustering from Spain which won first place for its solution to preserve the Mediterranean Sea by predicting variations in sea temperatures and using satellite images to identify polluted areas in the Mediterranean.

Team Cognition consisted of Tsakane Koko (pictured above), Hesme Cronje (Grade 12, Curro Heritage House), Humbulani Mudziwa (Grade 12, Curro Academy Soshanguve), Anamika Beethasi (Grade 11, Curro Waterfall) and Tahlia Bell (Grade 10, Curro Mossel Bay).

The South Africa team prepared a presentation with ideas to help trace and locate African Wild Dogs using AI. Curros said in a statement that in the early stages of the hackathon, the learners were given a tour on how the challenge will work as well as shown online demos to work with in order to get a feel for the capabilities of AI. The next step was to take data and apply machine learning code to it. The end goal was for all the girls taking part in the challenge to design the outline of a web service that uses AI. Team Cognition then made use of AI to search social media posts containing any geotags or hashtags relating to African wild dogs or their known residents.

Curro Academy Pretoria phase head Charlotte Jooste said the Team Cognition presentation focussed on AI methods to pick up any indications in the wild dogs’ behaviour that could link to illness or other threats as well as interventions. “This way the animals receive little human intervention and therefore live a more “natural” life. The team’s presentation also covered ensuring that wild dogs be protected from geological disasters or processes like droughts, floods, etc. as well as human activities,” Jooste said.

She added that the challenge gave the team a whole new perspective of “new jobs” out there and careers they might look into in the future.

Microsoft lead for Digital Transformation in Education Mark East commented in a LinkedIn post that watching the girls fully commit to outline an AI-based solution to a sustainability problem really showcased how entrenched the current generation is in environmental concerns.

“The more they grasped the full capabilities of AI to create a sustainable future, the more focused they were on coming up with a viable solution. Today’s visionaries are determined to change the world for the better, and they’re excited to explore how they can use AI to do it,” said East.

East added that one of the greatest highlights of the hackathon came at the very end, when some of the girls took the time to share their thoughts on the experience.

“Their emotions ranged from excitement, to gratitude, to genuine surprise at seeing so many women in the IT sector. Many of the girls expressed how excited they were to learn that AI will play such a vital role in the future, as they were interested in exploring career opportunities in this field. Those who had used the hackathon as a gateway into the amazing world of AI said that they were really glad they participated,” he said.

2ND QUARTER 2021 | SYNAPSE

15

NEWS

UCT’S FIRST-EVER

Google Research Scholar Program recipients

/ By Carla Bernardo /

Assoc Prof Amir Patel’s research will enable the measurement of animal motion in the wild at an unprecedented level

Two researchers from the University of Cape Town’s (UCT) Faculty of Engineering & the Built En-vironment (EBE) have become the institution’s first-ever recipients of search engine giant Google’s Research Scholar Program.

T

he two recipients are Associate Professor Amir Patel and Dr Mohohlo Tsoeu, both from the Department of Electrical Engineering and its newly formed African Robotics Unit.

They are also the only Africans in the programme’s 2021 cohort.

The programme aims to support early-career researchers who are pursuing research in fields relevant to Google. It provides “unrestricted gifts to support research at institutions … and is focused on funding world-class research”.

To be eligible for the programme, an

16

SYNAPSE | 2ND QUARTER 2021

entrant’s research must be in computer science and related fields. Associate Professor Patel has been accepted in the Machine Perception category and Dr Tsoeu in the Natural Language Processing category.

“It feels amazing to be recognised by Google, one of the largest tech companies in the world. It is also exciting and encouraging for us to be mentioned among some of the world’s top universities in the field of robotics and artificial intelligence,” said Patel.

Tsoeu said that he was “excited and honoured” to be one of only two researchers from an African university to receive this prestigious honour this year.

“It is a positive affirmation that our research is important and has the potential to have great impact, and that we have the intellectual capital to deliver,” he said.

Invaluable, unprecedented insight Patel was awarded for his research project, “WildPose: 3D animal biomechanics in the field using multi-sensor data fusion”.

The project aims to provide deeper insight into the abilities of the world’s greatest animal athletes (located in Africa), such as the cheetah, and into how they can “robustly

traverse through the unstructured world”; it will prove invaluable for legged robots if they are ever to leave the confines of the laboratory.

Patel explained that currently the biggest impediment to a holistic understanding of animal locomotion is measuring and modelling whole-body motion in the wild.

Therefore, his project proposes developing a deep learning-based motion-capture system (WildPose), which leverages complementary sensor data to remotely obtain high-speed, whole-body 3D animal kinematics in the field from a single view.

The WildPose system will enable videographers to capture biomechanical data from animals such as cheetahs and lions in the wild, using a single handheld device, and creating a new source for data collection.

“This research is important, as it will allow us to measure the motion of animals in the wild at an unprecedented level,” said Patel.

Furthermore, Patel believes the proposed system will be disruptive in the field of ecology by providing new insight into how climate change affects animal behaviour; in neuroscience (how animals such as the cheetah respond to the motion of prey during hunts); in evolutionary biology (how

NEWS

the cheetah – the only surviving species in the Acinonyx genus – evolved, compared to Africa’s other big cats); and in robotics (holistically understanding animal locomotion will enable the design of more agile robots).

On the importance of Google’s Research Scholar Program and how it will support his research, Patel said that it is vital for enabling emerging researchers to explore artificial intelligence and robotics in exciting new domains.

“I believe this award will help me further my goal of moving biomechanics beyond the confines of the laboratory,” he said.

language divide remains a bottleneck to enjoying full global connectedness,” said Tsoeu.

In addition to contributing to new knowledge and new language corpora, Tsoeu’s and his team’s work will also open doors to “amazing research ideas and innovation in the areas of machine learning, human–machine interfacing, media content distribution, and education”.

But all of this costs money, and a lot of it.

Therefore, the funding Tsoeu will receive

from the programme will make a significant contribution to the project’s “enormous costs”, which include his and his team’s research activities, student funding, travel costs, fieldwork and attending conferences.

“I am honoured to have received it,” said Tsoeu.

This article was first published on www.news.uct.ac.za by Carla Bernardo, with photos by Je’nine May.

Restoring dignity Tsoeu was awarded for his project, “Corpora collection and complete natural language processing of isiXhosa, Sesotho and South African Sign Language”.

His project will contribute to the development of comprehensive, high-quality language corpora for indigenous South African languages. It will also investigate and develop novel and high-performance machine learning algorithms aimed at application areas such as automatic speech recognition, translation and text-to-speech/ sign technology.

Tsoeu explained that these applications are in the growing area of human–machine interfacing. But more importantly in the South African context, they contribute towards bridging the human language divide and improving equal access and participation to restore the dignity of currently marginalised groups, such as the Deaf and hard-of-hearing communities.

“My passion is to do research that drives good social change,” he said.

“I believe that in conducting such research, one opens avenues for good fundamental and applied research questions that contribute new knowledge and advance scholarship.”

His research interests are in machine learning, with specific applications to aspects of natural language processing.

This includes language designing, language corpora for machine learning, investigating novel algorithms for language translation, automatic speech and sign recognition, and text-to-speech and sign synthesis. Tsoeu said that it is a multi-disciplinary research area that stimulates collaborations between engineering, linguistics and the psychology of learning, and as he puts it, “is really fascinating”.

Assoc Prof Amir Patel was awarded in the Machine Perception category.

Photo: Je’nine May

Bridging the divide Tsoeu’s research contributes to “bridging the human-to-human language divide that devastates South Africa, leading to political language debates at universities and other spaces, and marginalisation of native speakers of native languages, especially the Deaf community”.

“The world is getting extremely connected, both through travel and the web, and the

Dr Mohohlo Tsoeu is participating in the Natural Language Processing category.

Photo: Je’nine May

2ND QUARTER 2021 | SYNAPSE

17

FEATURE

CHANGING TIMES:

How the pandemic pivoted this SA engineering consultancy into a 4IR SaaS platform / By Daniel Mpala / South African industrial engineer Katlego Malatji believes that when it comes to manufacturing, Africa is lagging so far behind with technology that it is making the continent less competitive globally. “It’s really sad because we are one of the richest continents in terms of raw materials, but we are not doing a very good job in terms of processing that raw material into value added goods and then selling them,” she says.

S

he explains that, for example, South Africa’s manufacturing industry is operating at a comparative disadvantage relative to countries like China who are actively using Fourth Industrial Revolution (4IR) technologies like automation, robotics, AI and the Industrial Internet of Things (IIoT). She says as a result of the efficiencies realised by these technologies, China is selling goods at a fraction of the cost, while local manufacturers on the other hand are losing out — despite working twice as hard — as their processes are much more labour intensive.

Malatji — who’s over the past six years built a name for herself as an international 4IR implementation specialist working for automative manufacturer BMW — points out that globally, the manufacturing industry is generally “moving in the direction of 4IR”. She strongly believes this is the trajectory Africa should be taking too.

This need for African 4IR solutions is what inspired her to start her own consultancy — ProjectOne Engineering. Malatji studied Aerospace, Aeronautical and Astronautical Engineering at Wits University for three years ending in 2014, thereafter she took up an Industrial Engineering diploma at Tshwane University of Technology (TUT) before attaining her BTech in Industrial Engineering at the University of South Africa.

“My expectations going into studying engineering were that we were going to be

18

SYNAPSE | 2ND QUARTER 2021

working on 4IR. I didn’t know the name at the time, but I had the idea that engineering is finding the solutions for the future - the robotics of it, the system optimisation — but that wasn’t the reality of it, at least not then in South Africa,” she says.

Her first job saw her work as an Industrial Engineering trainee at a company in Brits which specialised in supplying electrical protection and factory automation solutions.

She remembers that during her stint there in 2015, she noticed that there was not much tracking of data going on at the company.

“Data wasn’t seen as gold, which already at the time I thought ‘We need data, we need much more data than what we have’,” she recalls.

Later that year Malatji joined BMW South Africa (SA) as an intern focusing on quality management, and particularly on optimising processes to minimise defects. She says she learnt a lot from that role. “It felt like I was at a a new university because they had automation —to a level, they had systems that were in place, standards that were in place, and data was gold there,” she explains. However, despite this, within six months Malatji began seeing opportunities for implementing 4IR techniques. Some of the suggestions she made included improvements like using cameras to detect defects. She says at the time the responses she got were always around “budgets and business cases”. “They had processed that had been going on for years and years, and those processes were working, they were good processes that were

working. My ambition was for there to be more new technology put into place, not just for the sake of technology, but for the sake of seeing us move from one level of production to the next,” she explains.

In 2017 Malatji made history in the company by becoming the first South African to be independently hired from a BMW SA plant to work at BMW US. In her new role as a Logistics Integrator at BMW Manufacturing in Greenville, South Carolina she chose a project which allowed her to work on the full-scale introduction of automated guided vehicles (AGVs) — something she said had never been done at any other BMW plant at the time. In essence the project entailed taking normalsized fork lifts and tuggers and outfitting them with hardware and software which effectively made them autonomous. “ At the time I didn’t know it, but when I got to the US it was apparent that it was a very big deal,” she says.

On the successful completion of that project, Malatji then applied to join another project with BMW UK, this time working on logistics material flow and supply chain. This was around peak Brexit, with the company trying to figure ways to optimise its supply chain — particularly routes and truck loads —in case the borders shut down. “There was a lot of innovation, there was space for innovation, space for free thinking,” she says.

The objective of the project Malatji worked on during her stint with BMW UK involved making £1-million in savings per year. Within six months the South African engineer had

FEATURE

managed to save the company £1.2-million through the use of route-optimising algorithms. Her managers, she says, told her that the company had had trouble with that particular objective “for quite some time”.

“I asked them what had been happening in the previous year, it was six months before I quit. Because immediately then I was like I’m ready to go start my company.”

Starting ProjectOne IoT In September 2019 Malatji moved back to South Africa and registered her consultancy.

Initially she was working with two contractors.

“I was the only full time employee. The first project that we got was a construction company that needed AI to optimise their material, the second one was a yoghurt production company that wanted to increase their production capacity — they had more orders than they were able to fill — so they needed to design a process that would help them optimise the resources that they had.”

“Then the third one that we had was a supply chain project, it was an online store that wanted to digitise the way that they were tracking their material and storing it. We had those projects, the first one from December 2019, and the others from February 2020 and then in March we were still talking to larger manufacturers about their processes.

We were talking to a mining company in Germiston that wanted to improve their bearings. But when COVID hit, all those projects just slowly came to a halt. Some of them are still talking about resuming, while some of them were paused indefinitely,” she says.

Malatji says when these projects came to a halt, she realised in May that the pandemic would affect her business. “How do you consult for a company that’s closed? This is three months after I’d started ProjectOne and already there was a global crisis.”

At the time Malatji says she had been working on AI platform that would optimise systems in manufacturing and she heard about an opportunity for South African startups with IBM. Malatji applied for the initiative — the IBM Tech Accelerator run in partnership with Tshimologong Digital Precinct— and was selected of the programme along with 13 other companies.

Malatji presented the AI platform to IBM and then spent three months talking to the company about it. “It was basically pitching to them but over incubation. They didn’t focus on ‘Oh you’re just pitching for us’, but they focused on how do you actually style your ideas and pitch for other people and grow from there.

“ProjectOne came out second and was awarded R300 000 towards the building of the platform and then they also partnered with ProjectOne to supply some of the software to the value of over R500 000.

From there I realised that given Covid19, manufacturers have to change their ap-

proach to how they do business,” she says.

She explains that the pandemic made manufactures see the importance of 4IR and the need to move away from traditional systems where you needed everyone on site for anything to happen. “They started to really see that there are processes that you can automate to keep your business going if anything happens. Companies who had that already, or an element of that already, did better than ones that didn’t have any automation at all. But overall, really when you talk to manufacturers now its a whole different conversation because 4IR used to be a nice to have - we are already doing so well in business, why change what’s working. Versus now, now they see it as a critical part of running business,” she adds.

The SaaS pivot She explains that the same way that manufacturers had to change, so too did her company. “We weren’t just consultants being able to go to a company one-by-one. I was able then to focus on the platform which would then change into a software-as-aservice (SaaS) product,” she says. Malatji says she’s still working with IBM on the SaaS platform, which is called ProjectOne IoT. “I have two go-to-market partners who are interested in offering the platform to their clients once it is ready. We are working on getting two manufacturers in the FMCG sector on board who can be proof of concept partners,” she points out.

ProjectOne’s approach, she explains, is a proactive one which relies on data analytics. “So we take big data processing and data analytics to identify the critical areas where waste, problems or defects might occur, and we are able to provide solutions before those costly problems or de-fects occur,” she says.

“A better example is where there are recurring defects. Recurring defects are a huge, huge source of loss. There are even defects that are accepted now because they (manufacturers) have done several projects on them, and it’s cost a lot, but there’s a control that you can put in there to say ‘Ok we put someone to inspect every single one of them,’ and when there’s a defect fix it on sight. Now we are saying, ‘Ok let’s go two steps back, the problem that you are seeing could even be another symptom.’

So the in-depth root cause analysis of an engineering problem using data is what puts the answers in the factory owner’s or managers hands.”

“Any sustainable solution should still be in use in the future. We are not saying we are going to solve every problem with 4IR, but if there’s a problem that we can solve, and there’s a 4IR solution that fits that problem, it makes more sense to go for that solution that is sustainable than to go for a temporary solution,” she says.

Malatji wants to launch ProjectOne IoT to the manufacturing industry in Q2. “It’s going to be part of an engineers toolkit — especially the consultants. It shortens the problem discovery part of their engineering project by using data and analytics. Because it’s a platform and it’s SaaS, we definitely want to go into the rest of Africa,” she says.

This 4IR approach to manufacturing, Malatji believes, will “definitely” put South Africa and the rest of the continent in a place where it will be globally competitive. “They (developed countries) are a little bit ahead of us, but we are all at the starting line of 4IR. Or at least of using AI specifically for manufacturing,” she adds.

Malatji however acknowledges that there’s a fear of AI and 4IR technologies in South Africa which she believes is driven by a misunderstanding of what these technologies and what they can do. She believes that there’s need for education and awareness on these 4IR technologies.

She explains that her company is focused on educating society on how 4IR technologies like AI do not actually exclude people from industry. The company recently held a webinar — in partnership with TUT’s Directorate of Research and Innovation and Pretoria North Regional Innovation Networking Platform — to introduce the benefits of 4IR in business.

Malatji explains that her company is — in conjunction with some academic institutions — working on an introductory course, which is likely to start in June, aimed at those who work in manufacturing. “It’s about the basic core steps of what goes on in 4IR. It answers the question why anyone should be interested in 4IR,” she says.

“ProjectOne is talking about the practical application of 4IR. We are talking about going on the ground and taking a system that used to be labour intensive and improving it,” she says. She adds that it’s also important to open up dialogue on 4IR with the government. “ It’s no use having a company like mine being afraid that people will say you’re taking our jobs and stopping the conversation there.

Government should also be backing up AI by bring excited about AI and 4IR and showcasing the technology so that we don’t have the same problem that we have now that there’s a shortage of skills in this area,” she adds.

2ND QUARTER 2021 | SYNAPSE

19

NEWS

NVIDIA UNVEILS ITS FIRST DATA CENTRE CPU US chipmaker NVIDIA in April unveiled NVIDIA Grace™ , its first data centre CPU that it says will deliver 10x the performance of today’s fastest servers on the most complex AI and high performance computing workloads.

NVIDIA Grace was announced by NVIDIA CEO and founder Jensen Huang on the first day of NVIDIA GTC 2021.

Availability is expected in the beginning of 2023.

NVIDIA said the Arm-based processor is the result of more than 10 000 engineering years of work and is designed to address the computing requirements for the world’s most advanced applications — including natural language processing, recommender systems, and AI supercomputing— that analyse enormous datasets requiring both ultra-fast compute performance and massive memory. NVIDIA Grace, the company pointed out, combines energyefficient Arm CPU cores with an innovative low-power memory subsystem to deliver high performance with great efficiency.

Huang pointed out that leading-edge AI and data science are pushing today’s computer architecture beyond its limits

20

SYNAPSE | 2ND QUARTER 2021

– processing unthinkable amounts of data. “Using licensed Arm IP, NVIDIA has designed Grace as a CPU specifically for giant-scale AI and HPC. Coupled with the GPU and DPU, Grace gives us the third foundational technology for computing, and the ability to re-architect the data centre to advance AI. NVIDIA is now a three-chip company,” he added.

Delivering Breakthrough Performance NVIDIA said Grace is a highly specialised processor targeting workloads such as training next-generation NLP models that have more than 1 trillion parameters. The company said when tightly coupled with NVIDIA GPUs, a Grace CPU-based system will deliver 10x faster performance than today’s state-of-the-art NVIDIA DGX™based systems, which run on x86 CPUs. Underlying Grace’s performance is fourth-generation NVIDIA NVLink® interconnect technology, which provides a record 900 GB/s connection between Grace and NVIDIA GPUs to enable 30x higher aggregate bandwidth compared to today’s leading servers.

Grace will also utilise an innovative LPDDR5x memory subsystem that will deliver twice the bandwidth and 10x better energy efficiency compared with DDR4 memory. In addition, the new architecture provides unified cache coherence with a single memory address space, combining system and HBM GPU memory to simplify programmability.

Grace will be supported by the NVIDIA HPC software development kit and the full suite of CUDA® and CUDA-X™ libraries, which accelerate more than 2,000 GPU applications, speeding discoveries for scientists and researchers working on the world’s most important challenges.

The company hopes Grace — named after US computer programming pioneer Grace Hopper, will serve a niche segment of computing. Some of its first users, Swiss National Supercomputing Centre (CSCS) and the U.S. Department of Energy’s Los Alamos National Laboratory, will use it to support national scientific research efforts.

NVIDIA said it is introducing Grace as the volume of data and size of AI models are growing exponentially. Today’s largest AI models include billions of parameters and are doubling every two-and-a-half months. Training them requires a new CPU that can be tightly coupled with a GPU to eliminate system bottlenecks.

The firm said it built Grace by leveraging the incredible flexibility of Arm’s data centre architecture. By introducing a new server-class CPU, NVIDIA is advancing the goal of technology diversity in AI and HPC communities, where choice is key to delivering the innovation needed to solve the world’s most pressing problems.

Arm CEO Simon Segars said NVIDIA’s introduction of the Grace data center CPU illustrates clearly how Arm’s licensing model enables an important invention, one that will further support the incredible work of AI researchers and scientists everywhere.

ADVERTORIAL

PUTTING AI INTO THE ENGINE ROOM unlocks real and instant value to large organisations with complex IT infrastructure

T

he real value of Artificial Intelligence (AI) to a business today exists in the engine room of Business Support Systems (BSS) and Operational Support Systems (OSS). As businesses drive their digitalisation strategies, they forget about the complexity that is created for operational management teams who must continue to monitor legacy systems while skilling up and supporting new systems. The operational environments are becoming exponentially more complex and existing staff are expected to support much more than before.

The typical answer to this emerging problem has been the drive towards DevOps structures where new systems are implemented and product subject matter experts provide both development and operational support functions. This allows for quick turnaround of incidents and “roll forward” of operational problems for that system, however, this does not solve the problem of end-to-end operational incident management and root cause analysis.

As a result, during service disruptions, emergency meetings are becoming large and complex as more people from different backgrounds are being pulled in to resolve problems. This process is generally managed by the IT Service Management (ITSM) operations who acts as a co-ordinating end-to-end team to manage the incident. Once the problem has been identified and fixed, a change is made to the system to prevent this situation from reoccurring, however, these fixes tend to be specific to one or two systems and it is post event. It does not prevent another system in the organisation from failing due to the same reason later as there is little to no learning shared across the organisational silos.

New systems being rolled out also lack operational support insights and therefore are prone to fail for the same reasons.

At Aizatron, our customers are some of the largest on the African continent. Their ICT systems are becoming increasingly more complex as new technology is required to deliver on the ever-growing business expansion. We leverage AI and Robotic Process Automation (RPA) technologies to automate smart operational bots that act across operational platforms and various company silos.

Our expert service offerings include: 1. Data Automation - Automated extraction of machine data from organisational systems as well as human generated data which is processed by our big data analytics engine.

2. Knowledge Acquisition - We work with existing staff and the ITSM knowledge bases to identify incidents and fixes to those incidents.

This is used to provide supervised learning to train operational bots algorithms which use past incidents as training data inputs.

3. Knowledge Transfer - Train existing organisational operations staff to develop their AI models to predict system outages and map out a course of remedial action. In this way, operational support teams become more effective in identifying and preventing outages while creating a self-learning organisation.

4. Automation Operations Manual - We maintain an operation manual (instructions) for the Bots like those that exist for humans, to ensure predictability and consistency in the operational

environment. The recommended course of action in the automation manual instructs the operational RPA bots on what actions to take when it is initiated by the AI module.

5. Multi-Bot Execution - The action can be taken by one bot or a set of bots – depending on the security policies of the organisation who has control of the bots. Some organisations create distinct divisions between systems for security reasons as well as to prevent employee fraud. These divisions need to be maintained at an RPA level as well so that organisational security is not compromised. In a multi-bot execution, one bot will complete its tasks and then send a message to the next bot to continue the process in its environment, following the same process humans would if they were executing the fix.

6. Customisation of Automated learning for specific situations The Bots have the advantage that they learn from the combined knowledge of the entire organisation, for example, lessons learnt from one system can automatically be applied to all systems and therefore, common problems do not repeat themselves. The machine learning identifies the root cause of the problem and based on the problem identification, chooses the appropriate fix to follow in the automation manual. These organisational learnings can be customised for specific scenarios by the system specialists where certain systems do not follow normal organisational best practices.

7. Actionable Operations - Automated actions can be taken by one or more bots depending on organisational security policies.

Organisations create distinct divisions between systems and resources enforcing segregation of duties between them for security reasons, to minimise employee fraud. These divisions need to be maintained and reported on to ensure organisational security and integrity.

Our approach to Smart operational management places AI technology inside the engine room of organisations by providing management with the essential dashboards to navigate the business more efficiently to drive bottom line profits.

Aizatron’s operational AI is used to identify the type of problem which it then uses to pick the right procedures to follow from the automation operations manual. Should it fail to resolve the problem, it will then escalate to a human to intervene. The operational AI will then be trained to determine the new scenario based on the data inputs and the human intervention will then be captured into the automation manual for future incidents.

As companies race to rollout their digitalisation strategies, they need to consider AI technology and smart bots to reduce their operational expenses and reduce the dependency on error prone humans. Humans tend to forget things or make mistakes and do not share critical knowledge, resulting in SLA breaches. Using AI to create smart RPA bots, ensures that learning is built into the organisational systems, thereby improving the business maturity levels by creating a learning organisation. It also frees up employees to focus on more complex, value added and creative tasks within the organisation, making it more agile, flexible, and adaptable to future emerging market and business conditions.

2ND QUARTER 2021 | SYNAPSE

21

NEWS

WILLIS RE LAUNCHES new South Africa Hail Catastrophe Risk Model

Reinsurance broker Wills Re in March launched a new Hail Catastrophe Risk Model which quantifies the risk from damaging hail events across South Africa.

T

he model, which was developed in collaboration with Willis Research Network Partners at NASA’s Langley Research Centre and the Karlsruhe Institute of Technology (KIT) in Germany, delivers a robust view of hail risk for the re(insurance) market.

Willis Re explained that the model features a comprehensive stochastic hail catalogue, developed by hail specialists at KIT using NASA satellitederived data. Applying the latest NASA hail detection algorithms, KIT mapped the distribution of observed frequency and severity. Events and their attributes were then simulated for a 25,000-year period to generate a comprehensive catalogue of hail events, with their footprints and parameter distributions reflecting observations in South Africa.

The company noted that the main applications of the Hail Catastrophe Risk Model outputs include: Pricing reinsurance contract layers for purchasing protection in the local and international reinsurance markets Independent sense-check for assessing capital adequacy and responding to regulatory solvency requirements Portfolio management and optimisation Willis Re pointed out that hail losses are relatively frequent in South Africa. “Seven of the top ten insured natural catastrophe events since the 1970s in the country are associated with hail, with upwards of 45% of the total value of insured Motor and Property claims from natural perils caused by hail damage. This frequency coupled with the potential for severe loss accumulations – as demonstrated by the catastrophic November 2013 losses in Pretoria – can also threaten insurers’ earnings potential,” the company stated.

Willis Re added that with this kind of prevalent risk at hand, a firm understanding of the potential financial impacts of a devastating 50-year hailstorm or the expected loss over the next ten years is critical.

Willis Re South Africa CEO and Head of Middle East and Africa Natalie van de Coolwijk commented that the updated hail model is an important addition to the firm’s toolkit and reaffirms its market leading position with respect to catastrophe modelling capabilities for the Middle East and Africa region. “ It will serve our clients in assessing their risk management needs for this prevalent peril and allow for a technical view in determining their risk appetite and mitigation measures,” added Van de Coolwijk.

Willis Re Head of Catastrophe Analytics for EMEA West-South MarieKristina Thomson said the firm is excited to launch its new hail risk model which will provide effective portfolio hail loss metrics enabling Willis Re clients to derive actionable insights. “In collaboration with our Willis Research Network partners, we have developed a robust view of hail risk

22

SYNAPSE | 2ND QUARTER 2021

across our territories in Europe and now South Africa, underpinned by the latest available data and detection techniques, empowering our clients to make confident risk transfer decisions for hail-sensitive portfolios,” said Thomson.

Kristopher Bedka, Research Scientist at NASA’s Langley Research Centre, said the NASA Applied Science Disasters Program — the sponsor of this NASA collaboration with Willis Re — KIT and a variety of international partners, promote the use of satellite observations to reduce risk and promote resilience from a variety of natural disasters including severe hailstorms on local and global scales.

“NASA satellite-derived products developed within this project are providing unique insights into severe storm climatologies over South Africa and other regions around the globe that have not been extensively observed by ground-based weather radar networks. We are excited to see that these data are providing significant contributions to risk assessment and development of this Catastrophe Model,” said Bedka.

Frequency of hail days across South Africa, taken from Willis Re’s South Africa Hail Model for Property and Motor classes (Willis Re)

“Willis Re explained that the model features

a comprehensive stochastic hail catalogue, developed by hail specialists at KIT using NASA satellite-derived data

”

ADVERTORIAL

5 STEPS

To Building A People Analytics Function From The Ground Up / By Elmen Lamprecht, Managing Partner, COGO People Analytics /

P

eople Analytics has taken the world by storm and many South African companies are now looking at implementing some form of advanced analytics in HR. The challenge is that most South African companies are too far removed from implementing predictive and prescriptive People Analytics. In fact, most companies have not even started on their People Analytics journey. This is not a surprise, since a LinkedIn study found that internationally only 29% of companies are using advanced analytics in HR. From our experience, we believe this number is even lower in South Africa.

At COGO People Analytics we help clients build a People Analytics function from the ground up. For them, artificial intelligence and advanced algorithms are just a bridge too far at this point. In this article, we provide 5 steps as a guide to building a People Analytics function from scratch.

Step 1: Discovery – Identify where People Analytics can make an impact in the Business HR is valued when it can drive organisational success through the optimisation of its people.

Just like Finance enables success through diligent management of money in the business, and IT drives productivity through the right hardware, network & software solutions, HR’s role in the business is to ensure that the people in the business contribute towards company goals. Therefore, the first step in building a People Analytics function from the ground up is to understand that priority should be given to organisational strategies and goals. People Analytics should not measure HR activities with no relevance to the larger business, but rather support the business strategy by measuring how people are contributing to what business deems important. Identify which People behaviours are aligned with strategies and goals.

Step 2: Examination – Selecting appropriate People Metrics Once you have identified where People Analytics will make the biggest business

impact, we need to determine which metrics we are going to use to measure this impact.

This is the heart of the process. At this stage, it is important to distinguish between HR Metrics and People Metrics. HR Metrics measures the Effectiveness & Efficiency of the HR function. Very often, this is totally separate from the business and has no relevance outside of the HR Department.

People Metrics measure the Effectiveness & Efficiency of the people in the business. This is where people behaviour shows a direct link to company performance. Your C-Level will only be interested in People Analytics.

Step 3: Data Mining – Obtaining Relevant Data Practically, the 3rd step is the most challenging. This is because the data you need for People Analytics is warehoused in several systems. To maximise this step, HR Professionals need to work with their ICT Department to develop a data management strategy that includes the following: Identify the sources of the relevant data (e.g. ERP Systems, stand-alone HR systems, email, Social Media, websites, engagement surveys – just to name a few).

Identify the type of data (e.g. Structured/ Unstructured, text/voice/image, GPS, video, etc) Harvest/gather the Data through system integration Normalise the data in a central database/ data lake Enrich data by adding of metadata (i.e. tagging)

Step 4: Assessment – Draw Insight from the Data Once we start receiving information through HR Analytics, we need to start making sense of everything. It is never enough to just take the data at face value. HR needs to interpret the data for the C-Level to clearly link the impact of People Behaviour on the bottom-line. We always propose looking at the following when interpreting data for the C-Level:

Provide Context – Current environment, Industry Benchmarks Compare to Past Performance – Progress/ Regression Provide understanding – Ask ‘Why?’ Predict future outcomes

Step 5: Influence – Communicate People Analytics to Drive Strategic Change The last step is often neglected. Positive, sustainable growth of People Analytics can only be achieved when HR is able to convince and influence the business by clearly communicating success stories to the business.

When communicating to the business, take the following into consideration: What should we communicate?

Structure message in a business-friendly story To whom must we communicate?

Identify all the stakeholders When should we communicate?

Weekly/Monthly/Quarterly Reporting Business Cycle (budgeting, Legislative Reports) How should we communicate? (Reports, Newsletters, Posters, Meetings, Emails, WhatsApp) By following these 5 steps, any organisation - regardless of size and sophistication – can build a People Analytics function from the ground up.

If you want to know more about building an HR Analytics function from scratch, please reach out to COGO People Analytics. We are Africa’s No 1 HR Analytics advisory.

We offer HR Tech and HR Analytics consultancy services, while our development of HR Analytics Tools (such as advanced HR Reporting which includes Artificial Intelligence and Machine Learning) and HR Virtual Assistants (Chatbots) remain our flagship products.

2ND QUARTER 2021 | SYNAPSE

23

ADVERTORIAL

FROM GARAGE TO GLOBAL:

How CompariSure’s conversational AI is driving digitisation within the Insurance industry

M

any of today’s tech giants, think Apple, Amazon and Google for example, had their humble beginnings in garage. For Cape Town based tech start-up CompariSure, the journey was no different. Perhaps it was the austere garage environment that led to creative thinking, or perhaps it was something else, but either way, it was one night in late 2018 while working from the garage when CompariSure Founder and CEO Jonathan Elcock (pictured, left) decided to code CompariSure’s first Facebook Messenger chatbot.

At the time, CompariSure was still reliant on the same “old school” digital technology that most players in the industry thought would bring a “digital revolution” a.k.a the traditional website. Fast forward to a little more than 2 years later, and over 1,000,000 people across South Africa have now interacted with CompariSure via its conversational commerce chatbot technology.

Even before COVID-19 plunged the world into crisis, artificial intelligence – and more specifically conversational AI – was fast becoming a key feature in the digitisation of customer-facing industries like insurance.

CompariSure’s focus on digital distribution

24

SYNAPSE | 2ND QUARTER 2021

via chatbots has allowed it to pioneer a new way of selling insurance in South Africa.

“In one experiment we ran, we found that one of our insurance partners was able to sell 25x more policies via our chatbot than via their website,” notes Monique Elliot, CompariSure’s Head of Marketing. “This for a traditional A/B test of course, controlling for all variables except the call to action of website versus chatbot,” add Elliot.

CompariSure’s founding vision was always to build a marketplace that connects users with quality insurance products from South Africa’s most trusted insurance providers. The profound realisation and insight along the way was that doing so via conversational AI would yield gamechanging results. The CompariSure conversational commerce platform allows companies to have personalised engagements with users at scale, driven by machine learning and data science.

“At the end of the day, conversation is a deeply human experience and we’ve found it to be more effective and able to deliver better customer experiences than any other form of marketing or channel of communication,” notes Matt Kloos, CompariSure co-founder and CFO (pictured, right).

Conversational commerce brings forth many benefits: consumers get instant

service; companies reduce costs; and human agents can spend their time solving more important issues that truly require the human touch. Despite the global industry moving in a digitised direction for some time, South Africa had been slow to adapt, resulting in an industry that is still heavily reliant on call centres to complete sales and provide customer service.

As the first and only authorised financial services provider (FSP) in South Africa to have successfully completed sales via a fully-automated Facebook Messenger chatbot, CompariSure is uniquely positioned to support the industry during this time of major digital transition.

CompariSure’s proprietary chatbot technology, which leverages popular platforms like Facebook Messenger and WhatsApp, has enabled heavy-weight traditional industry players like Sanlam, Old Mutual and Momentum Metropolitan to have automated, highly-personalised conversations with consumers at scale.

“Insurers are choosing to “future-proof” their business by partnering with CompariSure.”, notes CEO Elcock.

“After our breakthrough success in using our chatbot tech to distribute Funeral insurance products via our Marketplace, we started building white-label chatbots for insurers. Over the years, our team has perfected the art of partnering human empathy with machine learning to enable financial institutions to seamlessly integrate conversational AI into their service offering.”, says Kloos.

Underpinning the tech firm’s chatbot technology is a deep analysis of reams of data being generated via these authentic conversations. To date, the company has had 10s of millions of touchpoints with end-users, with every conversation point and path being analysed in a constant and ongoing effort to improve the customer experience. “By the end of the conversation, it appears that many end-users are genuinely unaware they were interacting with a chatbot. The chatbot often ends a conversation by wishing the user a good day further, to which many users reply Thanks so much - you too!“ chuckles Kloos.

“We put significant effort into ensuring the conversation flow is natural, to the point, and even fun and entertaining – the same way a top call centre agent would operate.”

CompariSure’s proprietary chatbot tech continues to evolve with the aim of bridging the financial and digital exclusion divide in South Africa and providing end-users access to a broad range of quality products in the most natural way possible – a simple “conversation”.

To learn more about

CompariSure and its tech offering, head to www.comparisure.co.za

A conversational commerce platform that allows companies to have personalized engagements with users at infinite scale, driven by machine learning & data science

www.comparisure.co.za

2ND QUARTER 2021 | SYNAPSE

25

Authorised FSP 48598

NEWS

Liquid Telecom rebrands to

LIQUID INTELLIGENT TECHNOLOGIES

P

an-African technology group Liquid Telecoms in March rebranded to Liquid Intelligent Technologies and announced its transformation from being a telecommunications and digital services provider to a full one-stop-shop technology group.

Liquid Intelligent Technologies explained in a statement that the rebrand highlights the group’s expansion of its cloud business, cyber security services, and other technologies added to its existing telecoms and connectivity capability. This, it said, furthers the group’s aim of accelerating

growth by providing tailor-made digital solutions to businesses in the public and private sectors across the continent.

As a Microsoft Gold Partner, Liquid Intelligent Technologies said it is redefining network, cloud and cyber Security offerings through strategic partnerships with leading global players, bringing innovative business applications, intelligent cloud services and world-class security to the African continent.

The group recently launched a cyber security business unit which the company said uniquely delivers security at its core, protecting business data throughout its lifecycle.

ISHANGO, AIMS

partnership to connect top African data scientists with international work experiences

I

shango, a London and Berlin-based social enterprise that creates highly skilled jobs in Africa the African Institute for Mathematical Sciences (AIMS) have announced a partnership to connect top African data scientists with international work experience opportunities.

Ishango said in a blog post in March that the partnership will boost African data science graduates’ employment prospects through a fully funded, three-month fellowship programme. The fellows, who will be hosted

26

SYNAPSE | 2ND QUARTER 2021

at the AIMS Rwanda Centre of Excellence in Kigali, will be equipped with technical and soft skills to prepare them for the international job market. Fellows will also get real-world experience and work remotely on projects where they will add value to international organisations.

Ishango co-founder Eunice Baguma Ball pointed out that leveraging data has become increasingly crucial for the growth of businesses and organisations across the globe. “Thanks to the work of institutions like AIMS, Africa now has a growing pool of bright and talented data scientists who have

Liquid Intelligent Technologies group CEO Nic Rudnick said the company’s ongoing investments in its networks and data centres across Africa have uniquely positioned it to use its infrastructure to accelerate the availability of new intelligent technologies including the high computing power of the cloud, artificial intelligence and cyber security to its customers. “We are now excited to be executing our vision of bringing new technological opportunities to the market with a highly differentiated product set supported by our existing infrastructure and digital innovation,” added Rudnick.

the potential to compete at a global level. At Ishango, our goal is to be the bridge between Africa’s top talent and global businesses.

Through this partnership with AIMS, we now have access to the very best that Africa has to offer. We look forward to working together to develop the data leaders of tomorrow,” she added.

AIMS Industry Initiative director Dr. Charles Lebon Mberi Kimpolo noted that with the growing technological advancements, things are becoming more and more connected.”In today’s smarter communities, traditional data processing software and techniques cannot deal with the analysis required to understand human behaviour. Specialised technical skills and tools are needed to deal with such large data and information. This is why we need to build capacity and create a critical mass of data scientists equipped with the required skills to understand how to perform complex data tasks across various businesses. AIMS is proud to partner with Ishango to leverage its partnerships and networks to provide AIMS alumni with opportunities to learn new skills and contribute to human capital development needed by partner companies. We believe this data science and AI fellowship program will boost AIMS alumni’s employability skills and enhance their transition to relevant employment,” he added.

“Ishango co-founder Eunice

Baguma Ball pointed out that leveraging data has become increasingly crucial for the growth of businesses and organisations across the globe

”

ADVERTORIAL

RPA: THE NEXT CHAPTER

In The Automation Story

/ By Dimitri Denissiouk, Managing Director, IBA South Africa /

R

obotic Process Automation (RPA) and Artificial intelligence (AI) are becoming more important for many organisations today, as the COVID19 accelerated the need to remove humans from a number of business processes to avoid the spread of the pandemic.

The benefits of improving workflow with RPA are not limited to the pandemic. These include a more efficient use of the time and potential of employees, improved customer experience, and a stronger control over business processes. Therefore, the focus is on saving time, improving quality, and reducing risk.

However, RPA can be a complex and expensive investment.

Total Cost of Ownership Managers always plan the Return on Investment (ROI) of any new project, but most managers are not pricing RPA correctly.

One should focus on the Total Cost of Ownership (TCO) rather than an immediate ROI. Naturally, an automation project results in faster processes, allowing the same team to be more productive, but additional longerterm benefits should be also considered.

First is the ability to transform your business. Many industries are experiencing a wave of rapid change. Expanding the scope of automation beyond what you can initially achieve is yet another significant advantage.

Managers need to understand that RPA is more like a platform on which other solutions can later be created.

EasyRPA The RPA world is largely monopolised by high priced licensed solutions and many businesses are faced with the need to automate, though they feel that automation is cost prohibitive or too complicated. It is much easier to venture on this journey, if you talk to someone that has already implemented real RPA solutions in real companies.

We believe that every company should have the opportunity to implement RPA. With this in mind and summarising many-year experience of intelligent automation, our experts have built EasyRPA, an RPA platform designed for development, deployment, running, and monitoring of software robots.

EasyRPA is not just a set of ready-made bots. It serves as a springboard to intelligent automation across the entire organisation.

Free License Our approach to intelligent automation is that an automation platform should come with a free license. In most cases, a company has to acquire an annual license for each bot in produc-tion. If we have 20 bots and the price is from $5,000 to $10,000 per bot per year, the investment is significant. Moreover, if you add a cognitive capability to your RPA endeavour, the cost will soar. This is why EasyRPA comes with a free license.

Citizen Developer or Professional Developer?

The goal of RPA is to streamline manual tasks across a number of applications and business units. The citizen model encourages business users to play a leading role in RPA. The pro-developer approach does not rule out the involvement of business users in RPA. It just encour-ages them to focus on identifying processes suitable for automation and on finding new opportunities for optimisation, while leaving bot development to IT professionals.

I would like to explain why we advocate for and use the pro-developer model. Technological complexities and legacies that reside inside organisations are difficult to manage. As RPA scales in complexity, it becomes hard to guarantee that the applications interact effectively or that the bots can navigate the interfaces as easily as humans can.

Unstructured data that constitute 80 percent of all organisations’ data make the situation even more complicated. Therefore, trained data scientists and AI engineers come into play to deal with intelligent automation solutions.

Center of Excellence Usually, clients build their centers of excellence to plan and implement RPA across their organisation. RPA, like any enterprise technology, requires input from developers, project managers, business analysts, and other IT staff. As an option, it is also possible to work with a consulting firm instead of building a large center of excellence in-house.

RPA Governance The right bot governance is another requirement. The most feasible way is to integrate the existing governance practices into the RPA development process. Libraries of reusable components must be a part of the core system capabilities.

Human-In-The-Loop RPA is often viewed as a technology that can replace people, but it is smarter to think in terms of how it can help people do their job better. For example, EasyRPA supports the human-in-the- loop feature, where human employees and RPA robots work together.

AI/ML Integration Finally yet importantly, an automation strategy should envision AI/Machine Learning integration. The RPA platform must enable developers to scale RPA initiatives and drive hyper-automation.

Scaling Innovation RPA is just one part of a transformation to a digital business environment. Strategies such as RPA can entirely redefine how a business model works. Technology changes quickly, but with a steady partner, you are prepared to head confidently into the future and open the next chapter in the automation story.

2ND QUARTER 2021 | SYNAPSE

27

NEWS

Aspiring data science students work on UmojaHack Africa 2021 problems at their universities across Africa (Zindi)

UMOJAHACK AFRICA 2021:

Over 1 000 students participate in Africa’s largest inter-university hackathon Over 1000 students from 126 universities across Africa participated in UmojaHack Africa 2021, a virtual machine learning hackathon hosted by data science competition platform Zindi on 27 - 28 March.

Z

indi said in a statement in April that more than $10 000 in prizes were awarded to data science students from 9 African countries, and more than 8500 submissions were made to solve three real-world machine learning challenges on Zindi.

Students from 21 African countries joined the event, representing Algeria, Benin, Cameroon, Côte d’Ivoire, Egypt, Ethiopia, Ghana, Guinea, Kenya, Malawi, Morocco, Nigeria, Rwanda, Senegal, South Africa, Sudan, Tanzania, Tunisia, Uganda, Zambia and Zimbabwe.

They participated in three different machine learning challenges: a financial resilience prediction challenge, a logistics challenge for African B2B service provider Sendy, and a computational biology challenge using the DeepChain™ platform developed by InstaDeep. The winning solutions developed by Zindi users will be shared with these organisations and deployed in real-world applications.

“Very happy with the outcome” In winning second place in the Sendy Delivery Rider Response Challenge, Tony Mipawa, a data science student from the University of Dodoma, Tanzania, epitomised the spirit of Zindi and of UmojaHack A year ago, Tony was a data science novice until he

28

SYNAPSE | 2ND QUARTER 2021

participated in Zindi’s first ever Mentorship Programme in 2020. He has grown in leaps and bounds since then, as evidenced by his prize-winning submission in this hackathon, less than a year later.

“I’m very happy with the outcome,” Mipawa said at the awards ceremony. “My advice is, whenever there is an opportunity to learn, you should take it. Learning is all about passion; whenever there is an opportunity to learn, put your whole effort into it, do it well.

Try to learn from anyone you meet. I would like to thank Zindi for what that mentorship programme gave me,” he added.

Global support UmojaHack Africa 2021 was sponsored by some leading names in the global and African tech, AI and financial sectors, including InstaDeep, Standard Bank Group, Microsoft, DeepMind, NVIDIA, and Old Mutual. They were integral in making the event a success by offering financial and professional development prizes, contributing their expertise and excitement to the event, and supporting UmojaHack Africa 2021 through their own channels.

Microsoft’s Principal Director for Software Partnerships Chris Lwanga said the firm was incredibly excited about this event spanning over 100 African universities and helping thousands of African students leverage their data science and AI skills to

solve African problems.

“At Microsoft we believe in empowering every organisation and person to do more,” added Lwanga.

“Standard Bank is deeply invested in funding and implementing critical data science skills development programmes, such as Zindi’s UmojaHack Africa 2021 hackathon, to position Africa as a serious competitor in the world’s rapidly emerging data-driven sector,” said Adrian Vermooten, Chief Innovation Officer, Standard Bank Group.

InstaDeeo CEO and co-founder Karim Beguir said his firm was delighted to support UmojaHack Africa again, describing it as “ an incredible initiative close to our hearts”.

“Seeing students from more than 120 universities come together to collaborate on real-world machine learning challenges is truly inspiring. This is, in our opinion, the best way to accelerate AI growth on the continent. Hackathons like UmojaHack bring us one step closer to achieving InstaDeep’s mission: building an AI-first world that benefits everyone,” he added.

Zindi CEO Celina Lee said UmojaHack Africa has proven to be a game-changing event, especially when so many young people have been impacted by the global pandemic. “This is a chance for students from across the continent to come together to learn, compete, and have fun. UmojaHack is about building skills, creating new machine learning applications to solve problems that really matter, while forging new connections among the students as well as with industry. We are incredibly excited to see what the students come up with in just one weekend,” said Lee.

NEWS

UNESCO

launches AI Needs Assessment Survey in Africa The United Nations Educational, Scientific and Cultural Organization (UNESCO) in early March announced the launch of an Artificial Intelligence (AI) Needs Assessment Survey in Africa, adding that it had organised a dialogue with African countries to reflect on the findings and the survey’s recommendations.

U

NESCO pointed out in statement on its website that its operational strategy for Priority Africa recognises the importance of knowledge for the sustainable socio-economic development of Africa and of capacity building in the field of ICT.

Artificial Intelligence, UNESCO said, presents a leapfrog potential for development in Africa, notwithstanding that these technologies are currently developed for and by companies, universities and governments outside of the continent. Countries can use

“Its findings will help guide

our discussions today on how Artificial Intelligence can be mobilised for the social good

”

AI and other digital technologies to catalyse their innovation ecosystems and accelerate sustainable development.

UNESCO Deputy Director-General and Assistant Director-General for Communication and Information a.i Xing Qu underlined the organisation’s ongoing efforts to support African countries to harness AI for their national development priorities. Stressing the significance of the publication, he noted, “Its findings will help guide our discussions today on how Artificial Intelligence can be mobilised for the social good”.

UNESCO said the survey highlights the need to strengthen policy, legal and regulatory knowledge for AI governance in Africa. Namibian ambassador to UNESCO Albertus Aochamub pointed out that the survey had captured the gaps in policy formulation across different domains like data protection, education and skills for AI among others. “At this day there is no

policy that governs AI in Namibia other than the 2007 Telecommunication Act,” added Aochamub. With only 51% of Namibia’s population having internet access, he described universal access to infrastructure as a challenge that needs urgent attention.

The survey notes that as AI policies are developed across Africa, countries will benefit from greater coordination and expertise to address similar and shared challenges. UNSESCO said Firmin Edouard Matoko, Assistant Director-General for Priority Africa and External Relations, who moderated discussions, noted how UNSESCO facilitates knowledge exchange among policymakers and other stakeholders in Africa through the organisation of several regional fora on Artificial Intelligence.

UNESCO said more than half the countries that responded to the survey identified gender equality related concerns with the use of AI as a priority.

UNESCO’s Assistant Director-General for Social and Human Sciences Gabriela Ramos pointed out gender equality is not only a question of women’s access to technology but also that of women as actors who shape technologies. “There is a need to have more women from Africa in shaping AI development as it is unacceptable that maleonly teams make 80% of the development in software linked to AI,” said Ramos.

Egyptian ambassador to UNESCO Alaaeldin Zakaria Youssef drew attention to how AI represents a significant economic advantage that can create new market opportunities. “AI offers the potential for a radical transformation of economic and social systems worldwide. But it also poses challenges. There is a need to follow up on the findings through international dialogue around new labour market paradigms, future of work and economic model redesign,” added Youssef.

UNSESCO said the survey findings will play an essential role in facilitating regional co-operation by acting as a starting point for countries to work together on shared priorities. Sally Radwan, representing the African Union Working Group on AI, shared the three objectives of the working group.

These consist of creating a common African stance on AI-related issues, identifying and implementing AI projects of mutual interest and establishing a framework for AI capacity building in Africa. She mentioned that the African Union’s Working Group on AI will discuss the survey findings at its next meeting.

The report can be read here, while the survey findings can be accessed here.

2ND QUARTER 2021 | SYNAPSE

29

NEWS

WITS, partners release AI-powered

ALGORITHM TO DETECT SA’S THIRD COVID-19 INFECTION WAVE The University of the Witwatersrand (Wits University) in April — in partnership with iThemba LABS, the Provincial Government of Gauteng and York University in Canada — released an AI-based early detection system which is capable of picking up South Africa’s third COVID-19 wave.

W

its explained that the system predicts future daily confirmed cases, based on historical data from South Africa’s past infection history, that includes features such as mobility indices, stringency indices and epidemiological parameters.

“These parameters are consistent with clinical public health measures that can contain, control and mitigate against the COVID-19 pandemic,” says Dr. James Orbinski, Director of the York University Dahdaleh Institute for Global Health Research.

Wits said the AI-based algorithm works in parallel, and supports the data of an already existing algorithm that is based on more classical analytics. Both of these algorithms work independently and are updated on a daily basis. The existence of two independent algorithms adds robustness to the predictive capacity of the algorithms. The data of the AI-based analysis is published on a website that is updated on a daily basis.

At the time of the announcement (11 April), Professor Bruce Mellado, Director of the Institute for Collider Particle Physics at Wits University said that while data from the model showed that the risk of a third infection wave was small across most of the provinces in the country, South Africa still remained highly vulnerable.

Wits pointed out that advent of infection

30

SYNAPSE | 2ND QUARTER 2021

waves is driven by circumstances that are difficult to predict and therefore to control.

In this complex environment, early detection algorithms can provide an early warning to policy makers and the population. Early detection algorithms are able to issue an alert when the data displays a significant change that is consistent with the advent of a new wave.

While algorithm-based predictions can never be 100% accurate, Mellado is confident that the model presents a very good prediction over at least a two-week period.

While predictions can be made over longerterm periods, these predictions become less accurate. The model is trained on the interim period in between waves one and two in all of the South African provinces. The algorithm was tested with data taken during the period of past peaks to evaluate its performance.

The project is supported by the Canadian International Development Research Centre (IDRC) through the Africa-Canada Artificial

Intelligence and Data Innovation Consortium.

AI technology provides us with invaluable potential to develop early detection and alert systems that are highly needed for rapid and dynamic decision making under risk and uncertainty under the current pandemic,” says Ali Asgary, Professor of Disaster & Emergency Management and Associate director of York University’s Advanced Disaster, Emergency, and Rapid-response Simulation.

AI, Wits said, is very effective in navigating through complex problems with a large number of parameters and dimensions, while at the same time learning from the data. Data hides within itself a wealth of information that AI can extract efficiently.

“Our team’s development of an early detection algorithm for the third wave speaks to the power of AI to generate data-based solutions to highly complex problems,” added Mellado.

NEWS

African startups Fastagger, Pyloop accepted into NVIDIA Inception programme Two African startups -- Kenya’s Fastagger and Nigeria’s Pyloop -- have been accepted into the NVIDIA Inception programme.

The NVIDIA Inception programme is a leading virtual accelerator run by American multinational technology company NVIDIA, with a focus on artificial intelligence (AI), data science and High Per-formance Computing (HPC) startups. The programme provides participants with critical go-to-market support, expertise, and technology.

Unlike traditional accelerators, NVIDIA Inception supports startups through their entire life cycle. There are no application deadlines, cohorts, or term limits.

Once a startup has joined NVIDIA Inception, they’re able to remain in the programme as long as they keep their membership active.

Nairobi-based Fastagger has developed an AI-as-a-service platform that provides data management, algorithms, and image annotation services to AI-driven companies. The startup was founded in 2019 by Mutembei Kariuki.

Port Harcourt based Pyloop uses AI and Earth Observation (satellite) data to monitor air pollution. The startup was founded last year by Gideon Onyewuenyi, Gift Kenneth, and Agwa Victor.

Coincidentally both Kariuki and Onyewuenyi were speakers at AI Expo Africa 2020 ONLINE. Check out Onyewuenyi’s talk here and Kariuki’s talk here.

SMART AFRICA,

Intel partner to build AI capacity building for African policymakers With AI set to increase global GDP by $15.7-trillion by 2030, there’s a need for policymakers to understand its benefits and risks to promote responsible AI that leads to sustainable economic growth.

Smart Africa -- an alliance of 31 African countries, international organisations and global private sector players tasked with the continent’s digital agenda -- in March through its capacity building arm the Smart Africa Digital Academy (SADA) and in partnership with Intel Corporation held a four-day capacity building workshop on artificial intelligence (AI) to empower African public sector decision makers on emerging technologies to drive informed policymaking, foster growth of the digital economy and promote national competitiveness.

The workshop, which was held between 15 to 19 March, is part of the Digital Readiness programme led by Intel which targets

policymakers responsible for designing, developing and deploying emerging technology-based solutions.

Government officials from 22 African countries participated in the four-day virtual event which focused on technology considerations for security in AI and the importance of data.

Smart Africa CEO and Director General Lacina Koné said the alliance has built a strong relationship with Intel Corporation over several years as an integral member of the Smart Africa Alliance. “This partnership is premised on the understanding that digital growth should be underpinned by strong and informed decision making. Capacity building for decision makers is a critical element through which we build towards a single digital market for Africa,” added Koné.

Intel Emerging Markets Director for Government and Education Sven Beckmann pointed out that there’s need for enabling policies that will unlock investments into the continent’s tech sector and facilitate skills development, mainly for Africa’s growing young population.

“Educating and empowering government leaders on emerging technologies to drive informed policymaking and expand digital readiness for all is of utmost importance.

We are excited to work with Smart Africa and look forward to our collaboration with government leaders for their digital transformation journey,” said Beckmann.

2ND QUARTER 2021 | SYNAPSE

31

NEWS

SANRAL EXPLORING MACHINE LEARNING APPLICATIONS

for road safety, congestion

“Some of the ways to mitigate

these potential privacy risks, are to use strict security and access controls. Data can also be anonymous at the point of capture. After all, the intention is not to observe individuals, but rather to identify trends and incidents to inform appropriate response and interventions

”

The South African National Road Agency (Sanral) in April announced that its Technical Innovation Hub (TIH) is looking into the use of machine learning (ML) for road safety, traffic congestion reduction and infrastructure development.

T

IH mechatronic engineer Ruan Van Breda explained in a statement how machine learning can be used to to detect and segment objects within a camera frame (each frame of a video is analysed as a still image). These objects, he added, can then be classified based on pre-trained image classifiers.

Van Breda said that the technology, when used within a road environment, had capabilities around the detection and classification of different types of vehicles, pedestrians, animals and cyclists. Sanral pointed out that there currently is already ample data available for these classification types.

32

SYNAPSE | 2ND QUARTER 2021

Van Breda explained that the creation of custom data sets and training classifiers would further expand the ability to distinguish between slow moving traffic and a road traffic crash. The agency said this can also be used to create new classification classes based on unique experiences like protest detection or for picking up foreign objects like rocks and tyres. This information, Sanral said, can then be used to activate the appropriate response through the Road Incident Management System (RIMS), remedy the situation and inform road users in real-time.

Van Breda further pointed out how the technology can be used to look at how different objects interact with one another, for example to detect unusual vehicle behaviour like stopping on the freeway.

“One is furthermore able to infer information about the interaction between multiple elements such as cars and pedestrians.

If a vehicle is detected moving to the side of the road and coming to a standstill and pedestrians are detected moving towards the vehicle and enter the vehicle, this can be classified as an informal pick-up. As more and more data is collected, these trends can either aid road authorities with infrastructure planning such as drop-off or pick-up points or aid law enforcement to stop illegal pick-ups if it is considered a

safety risk,” he added.

Sanral acknowledged how technology of this nature comes with significant risks.

The agency all efforts were being taken to understand how to effectively use the technology while maintaining strict compliance with legislation as it relates to the privacy of road users.

“Some of the ways to mitigate these potential privacy risks, are to use strict security and access controls. Data can also be anonymous at the point of capture.

After all, the intention is not to observe individuals, but rather to identify trends and incidents to inform appropriate response and interventions,” Sanral said.

“While this technology is still in the exploratory phase in South Africa, it already has tongues wagging in countries like China, where they use machine learning to incorporate facial recognition for law enforcement. They are able to identify the driver of a vehicle and instantly issue fines, if that driver does not have a valid driver’s license. Fines can also be issued automatically for individuals who jay-walk or gain access to restricted areas. As with any technological advances, there are pros and cons and in a complex society like South Africa, for now, let’s look and learn,” said Van Breda.

NEWS

Flickr via WOCinTech Chat, CC BY 2.0

STRATHMORE STUDY

lays bare lack of gender equality in African AI industry A recent study by the Centre of Intellectual Property and Information Technology Law (CIPIT) at Kenya’s Strathmore University looking into the representation of women in AI in Africa has found that women only constitute 29% of the continent’s AI workforce.

T

he gender disparity seems to be slightly higher relative to the global average which according to CIPIT stood at 78% men to 22% women in 2018.

The study titled, The Artificial Intelligence Labour Gender Gap In Africa, was conducted between last July and January this year.

CIPIT research fellow Dr Angeline Wairegi said in an email in April that the study is based on data gathered from 160 companies across 21 countries, namely Egypt, Morocco, Tunisia, Ethiopia, Kenya, Malawi, Rwanda, Tanzania, Uganda, Burkina Faso, Nigeria, Ghana, Sierra Leone, South Africa, Zambia, Zimbabwe, Cameroon, Algeria, Libya, Benin, and Mauritius.

CIPIT explained that the project mapped the gender composition of AI projects and companies originating in countries across Africa to capture the diversity struggles particular to AI startups, examine what those struggles exemplify in an African context, and determine the mechanisms that can be put in place to curb them.

For the purposes of the study, CIPIT defined African AI companies as those: Building application with machine learning and deep learning algorithms Located in Africa Operating at the time of data collection With at least two employees The study also analysed gender makeup of management across various departments.

Findings show women only make up 14% of founders, 13% of CEOs, 29% of CFOs, 38% of COOs, and 40% of directors, respectively.

“This study provides an analysis into the causes of this gender gap and their unique situatedness in Africa noting that, in part, there needs to be greater care into what AI is borrowed from the Global North and whether in borrowing tech we are not also

borrowing added layers of gender inequality to an already unequal system. The study also considers that an analysis of AI gender inequality in Africa requires a tripartite discussion nuanced with this study forming one part, the inclusion of women in the design of AI,” CIPIT stated on its website.

Gender parity in AI has been a growing topic in the industry because of concerns around gender bias in AI products. CIPIT stated that it wants to investigate the issue of gender bias in African AI systems and products in order to determine whether similar biases are present, as well as how African developers are looking to minimise such biases. Dr Wairegi said CIPIT hopes to submit the paper to a journal of choice in the next month or two.

Gender breakdown by Industry (CIPIT)

2ND QUARTER 2021 | SYNAPSE

33

INSIGHT

Does quantum computing hold the key to unlocking the secrets of the universe? To defeating climate change, helping humanity explore the stars or even preventing the next global pandemic? Could it help us accelerate the discovery of new drugs that combats incurable diseases, or help discover new materials, new ways of encryption or new weather prediction models? Perhaps it could help Wall Street simulate more accurate economic forecasts, optimise global portfolios and produce complex, in depth risk analysis.

34

SYNAPSE | 2ND QUARTER 2021

QUANTUM COMPUTING

could propel us light years into the future / By Rudeon Snell, Global Senior Director: Industries and Customer Advisory at SAP / Or is it destined to always be the technology of the future, never the present, similar to the promise of nuclear fusion?

The obscure and sometimes magicalsounding world of quantum computing is considered by some to be the most important computing technology of the century.

Amazon, Google, IBM and Microsoft, plus a host of smaller companies, such as D-Wave and Rigetti, are in a race to develop a commercially viable quantum computer.

China has also invested vast amounts of resources into developing their own quantum computing capabilities, with Origin Quantum pushing hard to catch up with leading global players in the field.

But what are quantum computers? How could they benefit humanity? And what realworld applications could we expect from this next-generation technology?

Crash course in quantum computing The computers we use today use bits, which are like tiny little switches that can either block or open the way for information to come through, essentially a binary structure of on and off. All data is made up of these bits, which are represented by ones and zeroes. And every photo, website, app and computer game is essentially made up of these ones and zeroes.

However, as our knowledge of physics has expanded, scientists have discovered that 'on' or 'off' isn't really how the universe works. In the natural world, most things are in a state of uncertainty, especially as you go down to a really small scale. Our present computers are not well equipped to deal with this level of uncertainty.

In contrast, quantum computers are designed to handle this uncertainty. It uses qubits, which can be 'on' or 'off', or both at the same

INSIGHT

time. Qubits allow computers to consider vast, previously unthinkable volumes of information, all at once. Instead of a normal computer considering each potential solution to a problem one-by-one, a quantum computer can consider all possible paths and outcomes at the same time.

This allows quantum computers to process information a lot faster, a lot more efficiently and at a scale previously unimagined. More importantly, however, because they use quantum mechanics, which are the foundation to physics and chemistry, quantum computers can analyse and efficiently process even the most complicated scenarios, something that remains outside the realm of possibility for even our most powerful supercomputers.

People often consider quantum computers to just be better computers, but quantum computers are as profoundly different to normal or even supercomputers, as a lightbulb is different to a candle.

Practical applications of quantum computing It's unlikely we will ever have quantum computing chips in our smartphones or home computers. Quantum computers are incredibly sensitive to interference, and have to be kept isolated and at nearzero temperatures (that's -273 degrees Celsius, colder than outer space). However, as researchers get ever closer to quantum supremacy - the point at which a quantum computer can regularly outperform normal computers - we are likely to see more widespread adoption for a broad range of use cases.

For example, in 2019, Google's quantum computer attempted a calculation that would

have taken our current supercomputers roughly 10 000 years to do. Their quantum computer completed the calculation in four minutes. This is the essence of what’s possible with a viable quantum computer.

Applications such as drug design and development, weather forecasting, climate change modelling and traffic optimisation could see quantum computing gaining mainstream use in the near future. For cybersecurity professionals, quantum cryptography holds the promise of total information security. While quantum computers could arguably be used to easily crack encryption codes thanks to their hyper-fast computing capabilities, using quantum encryption could make it absolutely impossible to break the code.

In financial services, quantum computing is expected to transform financial modelling.

JPMorgan Chase is part of IBM's Q Network, and aims to use quantum computing to improve risk modelling as a start.

Beyond risk assessment, practical applications of quantum computing in finance beckons to be explored, with researchers noting that "the entire financial market can be modelled as a quantum process". Simulating economic forecasts and optimising global portfolios are just a few of the untapped areas worth exploring.

Quantum computing is not only a next step in our technological development, it will accelerate the power and maturity of our current next-generation technologies, including artificial intelligence and advanced analytics.

Today's intelligent enterprises use cloud computing and intelligent technologies - such as AI and IoT - to accelerate productivity and enable real-time decisionmaking. With quantum computing, intelligent enterprises would be able to consider all possibilities at the same time, allowing decision-makers to manage any uncertainty in ways previously unimagined.

2ND QUARTER 2021 | SYNAPSE

35

NEWS

SIEMENS, CSIR MOU

to boost South Africa's 4IR German tech conglomerate Siemens and the Council for Scientific and Industrial Research (CSIR) on 3 February signed a Memorandum of Understanding (MoU) that will empower South Africa's economy and citizens with digital skills.

I

n a joint statement, the two organisations said the MoU will foster technical vocational education and training (TVET) around critical technical and digital skills to contribute to the employability of the local workforce and enhancement of the quality of job profiles.

The partnership will also see Siemens join the South Africa Centre for the Fourth Industrial Revo-lution (C4IR-SA) -- which is hosted at the CSIR -- and will assist in positioning the C4IR-SA as a thought leader in innovative digital technologies. The C4IR-SA aims to

36

SYNAPSE | 2ND QUARTER 2021

mobilise public-private partnerships to cocreate enabling governance framework which will optimally harness the potential of 4IR technologies -- such as artificial intelligence (AI), blockchain, and big data -- for societal development.

In addition, the partnership will also focus on piloting digital industry solutions in key sectors such as food and beverage, water, cyber security and manufacturing, as well as around smart and sustainable cities.

CSIR CEO Dr Thulani Dlamini (pictured above, left) , speaking at the signing event, said the partnership with Siemens forms part of the CSIR's strategy to foster relationships with the private and public sectors in order to respond to the needs of industry in order to improve the lives of South Africans.

"We are very pleased to join hands with Siemens in this huge and compelling task of ensuring that our country does not miss out on the gains of the 4IR. The CSIR strategy requires us to work very closely with the private sector to address the needs of industry and society, and to use science and technology to fast track digital skills of the future. To achieve this, the organisation is leveraging emerging technologies, especially those rooted in the 4IR, as well as its current capabilities and those of its partners,” said Dr Dlamini.

Siemens Southern and Eastern Africa CEO, Sabine Dall’Omo (pictured above, right) said the accelerated digitalisation caused by the coronavirus pandemic requires companies and society to respond faster and more

efficiently to changing market demands and in times of crises. “Siemens is proud to partner with the CSIR with this initiative and is ready to deliver on the fourth industrial revolution roadmap. Our goal as a company is to make sure that while we focus on continuously adapting, we’re also contributing to uplifting and building a sustainable economy,” said Dall’Omo. To align with the World Economic Forum’s (WEF) network of centres, Siemens has developed a comprehensive South Africa 4IR roadmap which will empower the country to seize the opportunities of digitalisation and especially Industry 4.0 solutions while up-skilling the South African workforce and creating new high-quality jobs.

Dr Dlamini said the 4IR has the potential to create high-quality employment opportunities across South African industries if South Africans are strategically skilled in futureoriented jobs. “Our partnership with Siemens will foster vocational education and training on critical technical and digital skills.”

“The business environment is getting more entrenched in the constant technological evolution and the industrial sector has been gradually integrating the use of automation and connectivity in its everyday business practices. This involves the digital transformation of industry to ensure that industrial processes become more adaptable, flexible and efficient and allows businesses to meet customer’s needs in the most reliable way,” said Dall’Omo.

NEWS

incorporating AI methodologies in their research activities. "Furthermore, in working together as opposed to working in isolated groups the impact of the research will also be improved,” says Barrett.

Wits said the establishment of Cirrus is therefore a rallying call to the academic and research institutions in Africa for there to be a critical mass of research and applications that can fully leverage the capabilities that will be established with Cirrus.

Dwolatzky said Wits has been on long journey with Cirrus to bring all of the elements of this ambitious partnership

WITS ANNOUNCES

team set to advance AI research in Africa Wits University has announced its team that will advance Africa’s AI initiative, Cirrus AI, a private sector led initiative bringing together academia and industry for the establishment of a world class artificial intelligence (AI) research and application capability for Africa.

T

he university team will be led by Professor Zeblon Vilakazi, ViceChancellor and Principal, with Professor Emeritus Barry Dwolatzky (project lead), Professor Nithaya Chetty, Dean of the Faculty of Science, as the champion for scientific engagement, and Dr Roy Forbes as the engagement coordinator.The university team will work closely with the Cirrus team that was announced at AI Expo Africa in 2020.

academic and research institutions to invigorate AI research and to further the application of AI across various academic and industrial domains.

Cirrus, which is the largest and most complex undertaking of its kind, is rallying the academic and research institutions in Africa to create a critical mass of research and applications that can fully leverage the capabilities that the initiative will establish.

Most of the institutions participating in the AIA Consortium will be publicly funded academic institutions from across the continent. With Wits University, Cirrus now has a leading university on the continent and a competent team to spearhead this important work.

South African and numerous other African universities currently host various academic and industrial research groups.

These groups are involved in, amongst other things, activities ranging from environmental and climate change research, medical research to materials research and energy storage development and design.

Cirrus founder Gregg Barrett points out that most, if not all, these existing research thrusts could benefit significantly from

into place. "We are hoping to soon begin to sign up members of the AI Africa (AIA) Consortium and to see tangible benefits flowing from our engagement with Cirrus,” added Dwolatzky.

Wits said its endeavours extend beyond the AIA Consortium and includes catalysing necessary strategic policy engagements with government to ensure impact on important research and societal objectives.

“The critical strategic steps that need to be taken in Africa have long been spoken about and now is the time for action. As Marc Andreessen recently pointed out, a takeaway from the COVID-19 pandemic is that people need to think about their occupation and contribution to society. If you are not helping people directly, and your occupation does not lead to something being built and contributes little to society, you are failing yourself. Cirrus represents Africa’s collective opportunity to move past the talk and get building on solving real problems with significant societal impacts,” said Barrett.

Wits said in a statement in March that its team is focused on two main priorities.

The first is to support the establishment of the AI Africa (AIA) Consortium which will provide the mechanism for bringing together academic and research institutions with a vested interest in the success and sustainability of Cirrus as Africa's AI effort.

The second priority is lead engagement and co-ordination with government on the adoption of Cirrus, support for local

2ND QUARTER 2021 | SYNAPSE

37

NEWS

SOUTH AFRICA’S CLEVVA

joins Blue Prism’s Digital Exchange (DX) Clevva, a South Africa web platform used to build, maintain and deploy front-office digital workers, has joined Blue Prism’s intelligent automation app store and online community Digital Exchange (DX).

By joining Blue Prism DX, Blue Prism said Clevva, along with two other companies which also joined store, would be making their software accessible to all.

UK-based Blue Prism, which is a global leader in intelligent automation for the enterprise, said in a statement in midApril that the latest capabilities on the DX enable organisations to take advantage of conversational AI applications, front-office automations as well as gaining insights from unstructured data such as emails, chat transcripts, outbound marketing materials, internal memos and legal documents, in a way that hasn’t been previously possible.

Blue Prism’s SVP Global Partner Strategy and Programs Linda Dotts said The Blue Prism DX community is a game changer because it enables, augments and extends our Digital Workforce capabilities with drag-and-drop ease of use. “It provides easy access to the latest innovations in intelligent automation through search and an a la carte menu of options. Our Technology Alliance Partners provide easy access to their software integrations via the DX, so everyone can drive better business outcomes via their Digital Workforce,” added Dotts.

Clevva helps customers realise straightthrough processing across staff-assisted and digital self-service channels. This solution

enables front office staff to navigate through rule-based decisions and actions, so they get it right, while driving intelligent self-service across any digital interface (website, mobile app, chatbot, social media, and in store kiosk). The combination of CLEVVA’s front office digital workers with Blue Prism allows customers to effectively automate end-toend processes across multiple channels in a consistent, compliant and context-relevant way.

Clevva founder and CEO Ryan Falkenberg explained that the company enables companies to capture the business logic that sits outside of operational systems—normally residing in experts, knowledge bases and decision tree scripts—and place it into a digital worker. “By coupling our ability to navigate customer engagements with Blue Prism’s ability to perform required system actions, we’re making end-to-end process automation a reality,” he added.

Clevva has been working with Blue Prism since 2016, this while the UK firm opened an

office in South Africa in 2020. In February, business publication ITWeb reported that the two firms had entered into a partnership that will see Clevva’s front-office digital experts combining with Blue Prism’s back-office digital workers to automate end-to-end processes across both staff-assisted and digital self-service channels.

Blue Prism VP Technology Alliance Programme Bruce Mazza told the publication the pairing of Clevva’s digital experts with Blue Prism’s digital workforce empowers users by bringing process agility to the front office.

“As organisations re-imagine customer sales and support journeys, many of them find it challenging to navigate contextually-rich or complex front-office decisions and processes.

This partnership will benefit workforces and customers − particularly in regulated industries, where customer engagements are heavily impacted by rules, and it’s vital you get it right, every time, with a record to prove it,” added Mazza.

INVESTMENT

NIGERIAN INSURTECH STARTUP CURACEL raises $450k pre-seed funding round L agos-based insurtech startup Curacel raised a $450 000 pre-seed funding round for its AI-powered platform which enables insurers to seamlessly automate claims, as well as track fraud, waste and abuse.

The round was led by by Atlantica Ventures and Consonance with participation from Kepple Ventures and other African angel investors.

The startup -- which was founded in 2017

by CEO Henry Mascot and CTO John Dada -- works with over 10 insurance companies

38

SYNAPSE | 2ND QUARTER 2021

across the continent, including AXA Mansard, Old Mutual, To-tal Health Trust, Hallmark HMO, and UAP. Curacel claims on its website that it has helped its clients reduce fraud, waste and abuse claims payouts by up to 25% and saved a total of $320 000, with over 700 000 claims processed.

Curacel will use the investment to expand to 10 new African countries by the end of the year. The startup is also looking to launch Curacel Capital, a cash advance product that makes it easier for healthcare providers to access working capital to mitigate financial challenges.

ADVERTORIAL

4 REASONS WHY

You Should Care About AI Governance NOW / By KOSA AI CEO & Co-founder Layla Li /

Have you seen AI bias make headlines recently from Amazon’s sexist AI recruiting tool to medical algorithms showing racial bias? In the new age of big data that we live in, this is increasingly relevant for companies that utilise AI’s automation power. If you’re not convinced why you should care, here are 4 reasons to act now.

1. AI Governance is becoming part of the regulatory landscape Maybe you think this is all too soon. AI is just a buzz word, your organisation is just starting a digital transformation, or you’re already market leader waiting to take the next step.

Regulatory changes are happening all over the world. According to the OECD, there are already over 300 AI policy initiatives from 60 countries and territories. Most notably, a leaked draft of the AI regulation by the European Commission reveals that the lawmakers are considering fines of up to 4% of global annual turnover or €20M (whichever is greater), for a set of risky AI use-cases.

Uber was recently sued by drivers in Europe over automated robo-firing. And Uber lost. The challenge references Article 22 of the European Union’s General Data Protection Regulation (GDPR) — which provides protection for individuals against purely automated decisions with a legal or significant impact.

If you haven’t yet started thinking about how you scale your AI operations responsibly,

2. Your customers care The risk mitigation factor pales in comparison of the benefits gained from strengthening your companies’ core values.

Companies that embrace responsible AI can make boost their bottom line and differentiate the brand, because your customers care.

Organisations with a strong sense of purpose are more than twice as likely to generate above-average shareholder returns, whereas AI without integrity will fail brands every time. Big players such as Salesforce, Microsoft, and Google have publicised the robust steps they have taken to implement Responsible AI. And for good reason: people weigh ethics three times more heavily than competence when assessing a company’s trustworthiness, according to Edelman research.

And more and more customers are consciously choosing to do business with companies whose demonstrated values are aligned with their own are. Companies that deliver positive impact on society boast higher margins and valuations.

Organisations must make sure that their AI initiatives are aligned with what they truly value and the positive impact they seek to make through their purpose.

3. Your employees care And not just your customers, your top talent also care about ethics. In UK, 1 in 6 elite AI workers has quit their job rather than help to build potentially harmful products.

Timnit Gebru, a leading AI Ethics researcher, was recently fired from Google after criticising its approach to minority hiring and the biases built into today’s artificial intelligence systems. And two engineers quit over the company’s treatment of their top talents.

A well-thought-out responsible AI program not only empowers talent to innovate with human impact at front and center of their work, but also can lead to improvements in recruiting and retention, who would have thought?

algorithms, or the teams responsible for managing them. This is not just a problem for gender inequality – it also undermines the usefulness of AI”.

Developing an algorithm that accurately performs on the whole spectrum of human diversity is also much more likely to deliver superior value to a broader and varied group of potential customers. Ultimately, expanding your AI program to ensure fairness and transparency will enhance revenue.

AI governance is a team effort. It’s not just up to your data scientists to figure out what safeguards to put in place against undesired outcomes from your AI systems.

Stakeholders across the organisations need to work together to develop a robust process that ensures accountability, transparency, fairness, safety and resilience. And always keep humans in the loop, be that your customers, your employees or the public.

4. More inclusive AIs = More revenues In the US, BCG research shows that companies lost one-third of revenue from affected customers in the year following a data misuse incident. Lack of trust carries a high financial cost. However, let’s not wait until disaster struck, the latent biases in your AI today are already costing you hidden millions.

Most AIs are trained on historic data that are coded with bias and missed opportunities. According to Gartner, “By 2022, 85% of AI projects will deliver erroneous outcomes due to bias in data,

At KOSA AI, we design and build Automated Responsible AI System that help multiple stakeholders in the organisation to evaluate and solve critical issues throughout the machine learning process, so they can understand and trust the AI systems they’re building.

2ND QUARTER 2021 | SYNAPSE

39

INSIGHT

MACHINE LEARNING SANDCASTLES: Considerations on why we need a dominant design & not another startup building a machine learning platform / By Gregg Barrett, Head of Cirrus /

A

t most organisations machine learning (ML) represents a disparate mix of tasks and tools, with data engineers working on data pipelines, data scientists on the data analysis, model training, validation and testing, hardware engineers on the compute configuration, and software engineers on the deployment. That tasks and tools are so segregated contributes to the high overall failure rates of machine learning projects and constrains the quality and quantity of ML project throughput.

This has led to the development of ML platforms. When I refer to an ML platform, I am referring to an all-in-one product for data and model development, scaling experiments across multiple machines, tracking and versioning models, deploying models, and monitoring performance.

The landscape for these all-in-one platforms is still underdeveloped though and while this might sound appealing for those with engineering and startup aspirations there is a need for caution.

Notes: Gartner: Data Science and Machine Learning (ML) Platforms

40

SYNAPSE | 2ND QUARTER 2021

INSIGHT The process from a data science standpoint The Cross Industry Standard Process for Data Mining, CRISP-DM, is an example of an open stand and process model that describes many common data science tasks that today typically form part of the ML lifecycle.

It is now increasingly common for organisations that have competence in ML to have a single person owning the entire lifecycle.

Treating ML as an engineering problem is a problem While it might be tempting to treat ML as an engineering problem those that do lose sight of the broad swath of users that are critical to the process and who are not engineers.

Subject matter experts for example frequently provide input on everything from data labeling and feature engineering to model evaluation. It is necessary for subject matter experts to effectively participate in the process while remembering that many are not programmatically savvy. For these users it is reasonable to envisage a low-code environment. To support these low-code users there is need to have layers of abstraction with each layer well defined. The highest level of abstraction — user interface (UI) driven ML — for most subject matter experts is yet to be realised, however for a moment imagine that we have the emergence of many ML platforms in the marketplace that provide a UI driven capability for these users. Without doubt each platform will operate differently. Pretty much everyone knows how to operate Microsoft Office because it is the dominant design — the design that dominates in the marketplace, and by large margin over similar offerings. In the case of Microsoft Office, it is the same design whether you are at home, at the office or at an academic institution. Without a dominant design there is no consistent and simplified user experience across organisations, domains, and tools.

Generic tasks (bold) and outputs (italic) of the CRISP-DM reference model.

Source

Distinct tones symbolise the distinct possible tasks for an individual or team to perform. A. Separating distinct types of tasks, specialists can address more conditions than one individual; the numbers add. What is shown is that two specialists can do twice as many tasks as one. For example, if each one can do 10,000 tasks, together they can do 20,000. B. Teams can address an even more diverse set of conditions because the numbers multiply. For a twomember team it would be 10,000 x 10,000 = 100,000,000. Source

Notes: Best Practices for ML Engineering

Rapid iteration where ML code is but a small piece ML platforms like general software systems need to be reliable, scalable and maintainable with the addition of being adaptable, as they are systems that learn from data — which can change frequently necessitating rapid development and deployment cycles. In addition, ML at present is in general an inherently empirical process with success in part proportional to experimental throughput, requiring rapid iteration. However, unlike traditional systems where programming is predominately the hard part, in ML programming is but a small fraction of the overall effort and dealing with the technical debt introduced from all the various components is a significant constraint on throughput.

Notes:Taxonomy of real faults in deep learning systems

2ND QUARTER 2021 | SYNAPSE

41

INSIGHT

Frameworks and API’s ML deployed in the wild can resemble a hydra of models and algorithms. A deployment could include, an ensemble of Principal Component Analysis, Gradient Boosting, and Neural Networks. In terms of an ML platform the decision on frameworks and libraries to support is being heavily influenced by Deep Learning although support for Deep Learning alone is wholly insufficient.

Supporting the predominant Deep Learning frameworks and libraries like PyTorch and TensorFlow, brings API’s into the picture. Given the nascency of the landscape with no overarching standard or dominant design, there is a plethora of API’s which are costly to develop and maintain. API’s need to be open sourced to ensure that architectures remain open and organisations are not forced into proprietary technology stacks. In addition, there is the need for development and access to data orchestration rules and APIs as a single interface in order to support the deployment ML across distributed environments (see ModelOps). This points to the need for a more universal platform and less diversification.

“While there is no dominant

design for an ML platform, open source has become the standard approach. There are a number of reasons for this and in the case of startups, organisations frequently require such so that in the event of the startup failing they have access to the source code

”

Rethinking architecture Rather than simply combining disparate systems that exist today to meet the unique requirements of ML systems, there is a need to rethink the design of the ecosystem supporting ML particularly when it comes to data — what is now termed a “data-centric” approach to ML. Whether to pursue a Data Orientated Architecture (DOA) or Microservice Architecture is an important discussion that needs to take place on this front. A DOA approach is essentially a streamingbased architecture, that makes data flowing between elements of business logic more explicit and accessible, thus the tasks of data discovery, collection and labeling for example are made simpler. Microservice Architecture is however pervasive due to its scalability.

42

SYNAPSE | 2ND QUARTER 2021

Notes: Modern Data Oriented Programming, Milan: An Evolution of DataOriented Programming & Data-Oriented Architecture

Open source as a standard While there is no dominant design for an ML platform, open source has become the standard approach. There are a number of reasons for this and in the case of startups, organisations frequently require such so that in the event of the startup failing they have access to the source code. Pursuing open source also means that the startup is now competing with existing open source tools and has to determine a viable business model that includes some sort of mix of proprietary and open source features. For a startup this is no easy task. If you are a large organisation like Google for example you can direct funds from profitable parts of the business to support open source efforts on TensorFlow to create an ecosystem with a virtuous circle of network effects — the more people use the framework, the more people know about it, in turn leading to more users — leading to a large sustaining ecosystem where there are opportunities to monetise proprietary tools and services like Google Cloud Platform. Google’s TensorFlow team is rumoured to be almost 1000 strong.

In terms of an ML platform I have to yet to see many opportunities for 1) establishing a large open source user community around a particular tool and 2) revenue opportunities from other sources that can fund the long runway to build such a community.

Notes: The Linux Foundation AI and Data, Mlflow, H2O

Focusing on a narrow use case Many startups are now attempting to develop ML platforms that serve a particular domain which amongst other things confines the startup to a smaller market and user base. My argument is that any ML platform must be agnostic to the use case as the core underlying technology is not domain specific. As the application of the underlying technology will not be confined to a single domain, the ability to build an enduring moat to prevent others, with a larger base because they are serving larger markets, from encroaching on a narrow use case will be found wanting.

Notes: Farewell to “Watson For Drug Discovery”

Focusing on a small step in the workflow An ML tool (I refer to “tool” in that it is not an all-in-one product) that only supports part of the ML lifecycle, say model training and evaluation, inevitably requires an organisation to stitch multiple tools

together. As products in the ML stack are constantly evolving and there is no common industry standard for interfaces, the cost of developing and maintaining the necessary integration across the ML workflow is nontrivial. Aside from all the integration headaches, the problems created for users having to be familiar with using multiple tools and user interfaces inhibits adoption.

For a large technology organisation where ML is a core component of products and services stitching together tools to create an ML platform is the current approach given the absence of a dominant design.

These organisations (unlike most) have the necessary skills, expertise, experience and resources to throw at the effort.

Such organisations will typically focus on interoperability to build an integrated solution spanning the entire workflow and for them it is sufficient that the platform handle only the use cases of the products and services the organisation is supporting.

Notes: Meet Michelangelo: Uber’s Machine Learning Platform, Productionizing ML with workflows at Twitter, TFX: A TensorFlow-Based Production-Scale Machine Learning Platform, Introducing FBLearner Flow: Facebook’s AI backbone

Tightly coupled components ML development and deployment environments are heterogeneous across organisations and an ML platform that is too tightly coupled with upstream and downstream software components will restrict its portability. Similarly, an ML platform that is too tightly coupled to a specific hardware accelerator will itself be restricted to the adoption of that hardware.

In such circumstances the startup needs to be careful to ensure that it is not betting on the adoption of a particular piece of hardware or software for the future of its market. An ML platform needs to work in as many environments, and on as many hardware configurations as possible which brings us to ModelOps and Domain Specific Architectures.

Enter ModelOps To avoid the manifestation of problems associated with hidden technical debt in production, Model Operations (ModelOps) is concerned with the best practices and tools used to test, deploy, manage, and monitor ML models in real-world production.

ModelOps is particularly relevant in managing the evolution of the model and data changes in the context of the underlying heterogeneous software and infrastructure stacks in operation across organisations.

With organisations moving to the cloud, the major cloud providers are now looking to integrate ModelOps with the rest of the

INSIGHT

organisational infrastructure which is driving a renewed focus on open integration across various ML tools and services. Such an approach by the major cloud providers is however easier said than done, as most have spent several years building proprietary walls around their products and services. Until such time as these platforms are truly open it is questionable as to whether any of these offerings will be the pathway to a dominant design for an ML platform.

context of ML also remains open. Building a language is a mountain of work and the speed at which ML is moving it would simply take way too long. Looking at what is in existence, the best options appear to be Julia or Swift. At present Swift has a little presence in the ML ecosystem and has mainly been used for iOS apps, however in recent years both Apple and Google have been moving it along in similar directions and from Googles side there is S4TF — Swift for TensorFlow.

Enter Domain Specific Architectures

Notes: Graphcore Poplar, Cerebras, Machine Learning Systems are Stuck in a Rut, Flashlight: Fast and flexible machine learning in C++

Domain Specific Architectures (DSAs), often called accelerators are a class of processors tailored for a specific domain.

This hardware-centric approach is driven by performance and efficiency gains as they are tailored to the needs of the application. Examples of DSA’s include Tensor Processing Units (TPU’s), and Graphics Processing Units (GPU’s). DSA’s use domain specific languages (DSL’s) to leverage memory access, parallelism and improve the mapping of the application to the domain specific processor. DSL’s are a challenge though as while being designed for specific architectures the software needs to be portable to different environments.

The vertical integration of the hardware/ software co-design for DSA’s is also supportive of open architectures as amongst other things this increases the number of users and improves security.

The question of what happens to programming languages like Python in the

Not understanding the requirement The absence of a dominant design for ML platforms results in many having a poor understanding of the ultimate requirement — the ML platform builders all too often simply don’t know what they don’t know.

This results in 1) organisations significantly underestimating the scope and complexity of the undertaking and 2) decisions being made (and justified) on questionable grounds.

Startups think that with a handful of engineers they can get something working.

Non-software technology companies in industry think it boils down to a build or buy. And academic institutions aided and abetted by external funding agencies delude themselves into thinking that developing an ML platform is somehow fundamentally

part of their research work and a good use or resources.

Ignorance, politics and fiefdom building aside, a major contributor to the build-it bias stems from people’s affection for their own creations — as crappy as it might be, those who built it ascribe more value to it. The reality is that most have grossly insufficient resources to achieve superior execution against the likes of Palantir, C3.ai, Databricks etc who have already had several years of runway and thrown a lot of resources at it — and who may not succeed at being profitable standalone businesses.

That for most organisations the value of ML lies in application and not from building and maintaining an ML platform seems obvious, yet the absence of a dominant design and poor understanding of the requirement leads many to make poor purchasing decisions. Ultimately this provides an opportunity for startups to provide professional services around specific organisational needs, however, there has been far less startup activity and traction in this area compared to those building tools and platforms.

Conclusion Ultimately the marketplace will settle the debate on any dominant design for an ML platform. Part of that process will be the evolution of financial conditions. Interest rates at a 4000-year nadir has misdirected capital, advanced speculation and perpetuated the unnatural lives of unsustainable businesses. When capital is eventually repriced, the demise of many large and small vendors in the ML marketplace, and the culmination of the eventual maturity of the industry will result in opportunities for consolidation and encourage the evolution of some sort of dominant design.

In the interim providers and users of ML platforms should think more strategically about the considerations raised to navigate towards an enduring value creating solution.

Additional resources: The Coming Wave of ML Systems, Stanford MLSys Seminar Series, Full Stack Deep Learning, Chip Huyen Automation. Source

2ND QUARTER 2021 | SYNAPSE

43

ADVERTORIAL

AUTOMATION ANYWHERE

Launches the World’s Only Unified Cloud-Native Platform for Intelligent Automation on the African Continent

Automation Anywhere, a global leader in robotic process automation (RPA), announced the introduction of Automation 360TM, a new brand for the company's unified, cloud-native, AI-powered enterprise automation platform. On top of this, customers on the continent can now deploy the unified platform in the AWS DC in Cape Town.

44

SYNAPSE | 2ND QUARTER 2021

A

dvancements to the previous Enterprise A2019 platform along with a new Automation 360 brand transforms the employee and user experience and delivers comprehensive process discovery, digitisation, automation and optimisation capabilities on a single, integrated platform to enable users to automate 2X more processes, at 3X the speed to scale an enterprise, and at a cost that is 1/5 of the infrastructure costs of legacy solutions.

The cloud has become the platform of choice for automation deployments, lessening the burden on IT resources and providing improved security, reliability, and flexibility as remote work remains in effect in many organisations. All on home soil.

"The leaders of organisations around the world today are seeking to

ADVERTORIAL

About Automation Anywhere

reinvent their enterprises for a post-COVID-19 era by using cloud automation to build more efficient, agile and resilient operating models," said Prince Kohli, CTO at Automation Anywhere.

"Our cloud-native Automation 360 platform clearly communicates to these leaders that if they want to fully empower people with a single, integrated, cloud solution for automating processes at scale across their entire enterprise, there is only one place to turn to, and that's Automation Anywhere."

Modern Automation That Transcends Legacy RPA The updated AI-powered cloud platform offers enterprises all the capabilities they need to build, deploy, manage and scale intelligent software bots, whether attended or unattended.

The result is an immediate improvement in employee engagement and customer experience.

Automation 360 allows enterprises to: Discover: The platform's Discovery Bot records user activities, documents business processes, and analyses process variances to identify which automation opportunities offer the highest business impact, and then generates bot blueprints that can be used to build software bots to automate these processes.

Automate: The platform's foundational intelligent automation product, RPA Workspace, includes significant enhancements to the low-code bot building experience, SaaS delivery models, and the centralised deployment, administration, and governance of bots. Citizen developers are equipped with more visual drag-and-drop experiences. RPA managers and IT teams benefit from the Cloud Preview Sandbox, with pre-release access to the next update to plan and develop ahead.

Digitise: Automation 360's IQ Bot has advanced its integrated Intelligent Document Processing (IDP) solution by combining RPA with pre-trained and custom-trained Artificial Intelligence (AI) models to automatically classify, extract and validate the information from business documents, emails and other unstructured

and semi-structured data, with minimal setup time.

Optimise: With Bot Insight, the platform analyses RPA and other bot activities in real-time, generating valuable business and operational insights for enterprises on how they can optimise their bots' performance at scale.

Meet AARI: Automation Anywhere Robotic Interface (AARI) is your digital assistant for work. Automate from anywhere. Increase RPA adoption with Citizen Development.

Be more productive and improve your office processes by automating repetitive, error-prone tasks across multiple systems.

Front Office Automation with AARI Desktop | Digital Assistant Demo

Cloud-Native Platform Advances Front Office and Back Office Automation According to a recent survey from Automation Anywhere, RPA is rapidly gaining ground in the front office. High volumes of users, paired with a surge in customer demand during the pandemic, attribute to the growing popularity of automation in customer-facing functions, such as call centres.

Customers are turning to a cloud-native platform, like Automation 360, which can be deployed in the cloud or as a hybrid solution that combines on-premises infrastructure with the cloud. The platform supports both legacy systems and modern applications, including support for SAP, Oracle, Workday, Salesforce, Office 365, G Suite and other enterprise applications– from the front office to the back office – enabling companies to extend end-toend automation across their organisation.

"Organisations are quickly realising that deploying their RPA initiatives in the cloud, and the productivity gains that come with it, offers increased efficiencies, reduced time to market, and improved customer and employee satisfaction," said Holly Muscolino, Research Vice President for IDC's Content Strategies and the Future of Work research services.

"In fact, IDC predicts that by 2022, 45% of repetitive work tasks in large enterprises will be automated."

Automation Anywhere is a global leader in Robotic Process Automation (RPA), empowering customers to automate end-toend business processes with intelligent software bots – AIpowered digital workers that perform repetitive and manual tasks, resulting in dramatic productivity gains, optimised customer experience and more engaged employees. The company offers the world's only cloud-native and web-based automation platform combining RPA, artificial intelligence, machine learning and analytics, yielding significantly lower TCO, higher security, and faster scalability than legacy monolithic platforms. Its Bot Store is the world's first and largest marketplace with more than 1,200 pre-built, intelligent automation solutions.

Automation Anywhere has deployed over 2.8 million bots to support some of the world's largest enterprises across all industries in more than 90 countries.

For additional information, visit www.automationanywhere.com.

1ai, part of the TreasuryONE Group of companies, is a certified preferred Automation Anywhere partner that design, implement and support RPA solutions in South Africa and Africa. With more than 20 years of experience in financial automation, your RPA project will be in the hand of experts. 1ai where we automate and innovate to save time, save costs and increase process compliance.

Visit www.1ai.co.za

2ND QUARTER 2021 | SYNAPSE

45

NEWS

AU, AFRICA CDC TURN TO 4D PARTNERSHIP

to tackle COVID-19 through AI, Big Data The African Union’s (AU) and Africa Centres for Disease Control and Prevention (Africa CDC) in May during a virtual meeting on initiatives to tackle the COVID-19 pandemic unveiled the 4D Partnership.

I

n a joint statement, the AU and Africa CDC described the 4D Partnership as a platform for multi-disciplinary, multi-dimensional, multi-departmental and multi-directional collaboration in Africa, powered by artificial intelligence (AI) and machine learning (ML), big data and home-grown innovation.

The 4D Partnership is a collaboration of several AU organs including the Commission’s Department of Education, Science, Technology & Innovation; the Department of Agriculture, Rural Development, Blue Economy & Sustainable Environment; the Department of Economic Development, Trade, Industry & Mining; the African Continental Free Trade Area (AfCFTA) Secretariat; and the Africa CDC. The platform will strengthen harmonisation of use cases in health,

46

SYNAPSE | 2ND QUARTER 2021

research, education, climate response and trade, for example, the Africa CDC’s Trusted Travel and Trusted Vaccines platforms.

Other programs, such the Africa Pathogen Genomics Initiative (Africa PGI), will benefit from the multi-stakeholder networks of the AU Science and Technology Framework, and help to accelerate integrated diagnostic, socioeconomic, immunological and genomic data streams to better analyse patterns of disease mutations and variants spread in Africa.

The 4D partnership will also speed up the emergence of electronic exchanges in the areas of climate finance, municipal green bonds, common fisheries rights, among others, enriching the quality of the postpandemic recovery already underway in parts of Africa.

The coalition behind the 4D partnership includes AfroChampions, the African Academy of Sciences, the African Institute of Mathematical Sciences, the PanaBIOS Consortium, the African Organisation of Standards, Koldchain BioCordon and the United Nations Development Program.

The Chair of the African Union, H.E. Félix Antoine Tshisekedi Tshilombo, President of the Democratic Republic of Congo, said: “The 4D Partnership is an early dividend of the recent reforms of the African Union Commission to better synergise its offerings and deliver clear impact for member states; and by leveraging the response to the pandemic to push innovation, it

is going to be a critical part of how we build resilience in Africa against future catastrophes.”

The Secretary General of the AfCFTA, H.E. Wamkele Mene, said: “Integrating a continent as vast, diverse, and rich in heritage, as Africa requires an ‘every tool in the toolkit’ approach yet resources are limited, hence the need to transform action-taking in Africa using the tools the 4D platform shall make available to key stakeholders across the continent.”

Africa CDC director Dr John Nkengasong pointed out that the pandemic had proven “quiet resilient to our usual responses”. “It is clearly time for a complete step-change in how we think about the recovery. 4D is how Africa’s continental leadership demonstrates its equality to the task,” he added.

Assistant Administrator and Director of the UNDP Regional Bureau for Africa, Ms. Ahunna Eziakonwa, said: “We take the opportunity given us to support Africa’s home-grown solutions and strategies very seriously and with a deep sense of commitment. As UNDP continues to play its technical lead role on the socioeconomic response to the COVID-19 pandemic, we will deepen our partnership with the AU and other key stakeholders, to advance transformative digital solutions that allow Africa to move beyond recovery, towards the 2030, sustainable development agenda.”

NEWS

Alphabet’s DeepMind establishes

SCHOLARSHIP FOR FOUR WITS MASTERS STUDENTS IN ML

British artificial intelligence research laboratory and Alphabet subsidiary DeepMind in May donated scholarship funding for four students from Wits University to complete their Masters degrees in machine learning.

W

its said in a statement that the scholarship will fund students registering for MSc degrees in Computer Science, Artificial Intelligence, or Robotics in the academic years between 2022 and 2024.

Wits added that the scholarships — known as the DeepMind Scholarships — will only be awarded to students who would not be able to take up their studies without financial assistance. Preference will also be afforded to South African citizens from underrepresented groups, including black students and women. DeepMind Scholarships will also be open to international students, with a preference to residents of Sub-Saharan African states.

The Scholarships will provide tuition fees, a stipend, plus conference and equipment funding for two Masters students who aim to complete their degree through dissertation over two years, as well as for two students who enrol in a Master’s programme through coursework and dissertation over two years. In addition, scholars get guidance and support from a personal DeepMind mentor.

“The spirit of the donation and the DeepMind Scholarships is to increase diversity in the fields of artificial intelligence and machine

learning, and to increase the representation of the groups currently most underrepresented in these fields,” said Obum Ekeke, Global Lead, University Relations & Education Partnerships at DeepMind.

“We are proud to help support the next generation of AI researchers and engineers in Africa.”

By participating in the DeepMind scholarships programme, Wits joins world-leading universities in the field of machine learning and artificial intelligence such as the University of Oxford, the University of Cambridge, New York University and University College London.

The university is one of only three African universities selected to host DeepMind Scholarships, alongside Stellenbosch University in South Africa and Makerere University in Uganda.

“Artificial Intelligence is an important building block and key driver in the Wits Digital Transformation suite of centenary projects, of which artificial intelligence and machine learning is a key driver,” said Professor Zeblon Vilakazi, Vice-Chancellor and Principal of Wits University.

Professor Benjamin Rosman, associate professor in the School of Computer Science and Applied Mathematics at Wits said the scholarships will become a much-needed platform and launchpad to the careers of some of the country’s most talented students in Artificial Intelligence and Machine Learning at Wits.

“This is an exciting recognition of Wits’ role as a leader in Machine Learning and AI in Africa. Enabled by this support from DeepMind, Wits will bring an even broader range of African talent to the global conversation in cutting-edge AI research,” said Rosman.

Prospective Masters students in the fields of AI and machine learning are encouraged to visit the Wits website for more information on how to apply for the DeepMind scholarships.

2ND QUARTER 2021 | SYNAPSE

47

INSIGHT

AI GONE GLOBAL: Why 20,000+ Developers from Emerging Markets Signed Up for GTC / By Kate Kallot, Head of Emerging Areas, NVIDIA /

GTC drew more than 20,000 registrants from 95 countries in Africa, Latin America and the Middle East

M

ajor tech conferences are typically hosted in highly industrialised countries. But the appetite for AI and data science resources spans the globe — with an estimated 3 million developers in emerging markets.

Our recent GPU Technology Conference — virtual, free to register, and featuring 24/7 content — for the first time featured a dedicated track on AI in emerging markets.

The conference attracted a record 20,000+ developers, industry leaders, policymakers and researchers in emerging markets across 95 countries.

These registrations accounted for around 10 percent of all signups for GTC. We saw a 6x jump from last spring’s GTC in registrations from Latin America, a 10x boost in registrations from the Middle East and a nearly 30x jump in registrations from African countries.

Nigeria alone accounted for more than 1,300 signups, and developers from 30 countries in Latin America and the Caribbean registered for the conference.

These attendees weren’t simply absorbing high-level content — they were leading conversations.

Dozens of startup founders from emerging markets shared their innovations.

Community leaders, major tech companies and nonprofits discussed their work to build resources for developers in the Caribbean, Latin America and Africa. And hands-on labs, training and networking sessions offered opportunities for attendees to boost their skills and ask questions of AI experts.

48

SYNAPSE | 2ND QUARTER 2021

We’re still growing our emerging markets initiatives to better connect with developers worldwide. As we do so, we’ll incorporate three key takeaways from this GTC:

1. Remove Barriers to Access While in-person AI conferences typically draw attendees from around the world, these opportunities aren’t equally accessible to developers from every region.

Though Africa has the world’s fastestgrowing community of AI developers, visa challenges have in recent years prevented some African researchers from attending AI conferences in the U.S. and Canada. And the cost of conference registrations, flights and hotel accommodations in major tech hubs can be prohibitive for many, even at discounted rates.

By making GTC21 virtual and free to register, we were able to welcome thousands of attendees and presenters from countries including Kenya, Zimbabwe, Trinidad and Tobago, Ghana and In-donesia.

2. Spotlight Region-Specific Challenges, Successes Opening access is just the first step. A developer from Nigeria faces different challenges than one in Norway, so global representation in conference speakers can help provide a diversity of perspectives.

Relevant content that’s localised by topic or language can help cater to the unique needs of a specific audience and market.

The Emerging Markets Pavilion at GTC, hosted by NVIDIA Inception, our acceleration

platform for AI startups, featured companies developing augmented reality apps for cultural tourism in Tunisia, smart video analytics in Lebanon and data science tools in Mexico, to name a few examples.

Several panel discussions brought together public sector reps, United Nations leads, community leaders and developer advocates from NVIDIA, Google, Amazon Web Services and other companies for discussions on how to bolster AI ecosystems around the world. And a session on AI in Africa focused on ways to further AI and data science education for a community that mostly learns through non-traditional pathways.

3. Foster Opportunities to Learn and Connect Developer groups in emerging markets are growing rapidly, with many building skills through online courses or community forums, rather than relying on traditional degree programs. One way we’re supporting this is by sponsoring AI hackathons in Africa with Zindi, an online forum that brings together thousands of developers to solve challenges for companies and governments across the continent.

The NVIDIA Developer Program includes tens of thousands of members from emerging markets — but there are hundreds of thousands more developers in these regions poised to take advantage of AI and accelerated applications to power their work.

This article was first published on NVIDIA’s blog on 4 May 2021

NEWS

ITU, AI Media Group launch the AI for Good

INNOVATION FACTORY (AFRICA) CHALLENGE The International Telecommunication Union (ITU) has partnered with the AI Media Group and AI Centre of Excellence (AICE) to launch The AI for Good Innovation Factory Africa challenge focused on the application of AI for Good to address SDGs and part of a wider global movement through the AI for Good Global Summit.

T

he AI for Good Innovation Factory aims to help innovators with scalable AI ideas as well as start-ups & scale-ups who are working towards achieving the SDGs by providing them with the opportunity to pitch their projects at the AI for Good Global Summit, a unique international and inclusive platform of AI innovators, sustainable development experts, UN partners, and public and private sector leaders.

This will also be the platform for the AI for Good Global Summit partners to design, announce or launch global challenges. These

challenges can be of different forms and formats, depending on the challenge nature and the implementing partner. It varies from an online crowd data challenge on a specific topic, to a moonshot grand challenges with global impact that takes a few years to achieve.

The winner of the Africafocused challenge will also be showcased at AI Expo Africa 2021 ONLINE in September

2ND QUARTER 2021 | SYNAPSE

49

INDUSTRY

IBM WANTS TO ACCELERATE DIGITAL TRANSFORMATION

with these breakthroughs in Hybrid Cloud, AI Capabilities IBM in May, at its 2021 Think Conference, unveiled advances in artificial intelligence (AI), hybrid cloud, and quantum computing that it believes will help its clients and partners accelerate their digital transformations, return to work smarter, and build strategic ecosystems that can drive better business outcomes.

I

BM Chairman and CEO Arvind Krishna said in a statement that we will look back to 2020 and 2021 as “the moment the world entered the digital century in full force”. “In the same way that we electrified factories and machines in the past century, we will use hybrid cloud to infuse AI into software and systems in the 21st century. And one thing is certain: this is a future that must be built on a foundation of deep industry collaboration. No one understands this better than IBM, which is one of the reasons we are boosting investment in our partner ecosystem. Also at Think 2021, we are unveiling our latest hybrid cloud and AI innovations – the very technologies that serve as the building blocks of a new IT architecture for business,” added Krishna.

The firm further stated that it is “all-in” on hybrid cloud and AI because it understands that businesses need a clear and credible path to modernising their mission-critical systems. IBM pointed to a new study that it had conducted on the adoption of AI for business which revealed that the imperative to embed AI into business

50

SYNAPSE | 2ND QUARTER 2021

processes became more urgent during the pandemic. IBM said 43% of the IT professional s surveyed in the study said that their companies had accelerated their rollout of AI, this while nearly half of global IT professionals surveyed said they evaluate AI providers in large part on their ability to automate processes. These findings, IBM said, are why it has invested heavily in building rich and powerful AI capabilities for business.

Here are some innovations IBM unveiled at 2021 Think Conference: Cloud Park for Data: A breakthrough capability in Cloud Pak for Data that uses AI to help customers get answers to distributed queries as much as eight times faster than previously and at nearly half the cost of other compared data warehouses. AutoSQL (Structured Query Language) automates how customers access, integrate and manage data without ever having to move it, regardless of where the data resides or how it is stored. AutoSQL solves one of the most critical pain points customers are facing as they look to reduce the complexity of curating data for AI and eliminate the high cost of moving data, while also uncovering hidden insights to make more accurate AI-driven predictions.

Watson Orchestrate: Watson Orchestrate is a new interactive AI capability designed to increase the personal productivity of business professionals across sales, human resources, operations and more. Requiring no IT skills to use, Watson Orchestrate enables professionals to initiate work in a very human way, using collaboration tools such as Slack and email in natural language. It also connects to popular business applications like Salesforce, SAP and Work-day®. Watson Orchestrate uses a powerful AI engine that automatically selects and sequences the pre-packaged skills needed to perform a task, and connects with applications, tools, data and history on-the-fly. This can help workers more quickly perform routine tasks, such as scheduling meetings or procuring approvals, or more mission-critical tasks, like preparing proposals or business plans.

Maximo Mobile: An easy-to-deploy mobile platform with IBM's leading Maximo asset management solution at the core, Maximo

INDUSTRY

Mobile is designed to transform the work of field technicians who maintain physical assets such as roads, bridges, production lines, power plants, refineries and more. A new, intuitive interface provides technicians with the right asset operational data at the right time. Even in the most remote locations, users can access Watson AI and in-depth organisational knowledge to easily solve complex issues. This powerful combination of AI, intelligent workflows, remote human assistance and access to digital twins puts decades of industry experience directly into the hands of technicians for safer, more efficient operations.

Project CodeNet Dataset: IBM Research is releasing Project CodeNet, a large-scale, open-source dataset comprised of 14 million code samples, 500 million lines of code and 55 programming languages, to enable AI's understanding and translation of code. Project CodeNet is currently the largest, most differentiated dataset in its class and addresses three main use cases in coding today: code search (automatically translating one code into another, including legacy languages like COBOL); code similarity (identifying overlaps and similarities among different codes); and code constraints (customising constraints based on a developer's specific needs and parameters). IBM believes Project CodeNet will serve as a valuable benchmark dataset for source-to-source translation and transitioning legacy codebases to modern code languages, helping businesses speed up their application of AI. Mono2Micro: IBM has added a new capability into WebSphere Hybrid Edition that enables enterprises to optimise and modernise their applications for hybrid cloud. IBM Mono2Micro uses AI developed by IBM Research to analyse large enterprise applications and provide recommendations on how to best adapt them for the move to cloud. It can simplify and speed up an errorprone process, which can reduce costs and maximise ROI. IBM Mono2Micro is one of IBM's suite of AI-powered products and services that can make it faster to migrate to the cloud Qiskit Runtime: IBM is making it faster and easier for developers to use quantum software by introducing Qiskit Runtime. This software is containerised and hosted in the hybrid cloud, instead of running most of its code on the user's computer. Together with improvements in both the software and processor performance, this allows Qiskit Runtime to boost the speeds of quantum circuits, the building blocks of quantum algorithms, by 120 times.

Qiskit, the IBM-developed open-source framework for quantum

computing for a global community of developers, aims to make quantum computing accessible to all. By introducing Qiskit Runtime, IBM is enabling quantum systems to run complex calculations such as chemical modeling and financial risk analysis in hours, instead of several weeks.

As part of a $1-billion investment to support its partner ecosystem, IBM unveiled new competencies, skills training, and benefits to ensure its partners succeed in an increasingly competitive market. For example, IBM has created a new competency framework to enable partners to demonstrate expertise, technical validation, and sales success in specialised areas such as hybrid cloud infrastructure, automation and security. IBM ecosystem partner Tata Consultancy Services (TCS) has already achieved competencies for building an industrial and manufacturing AI solution for data scientists and AI developers.

To further its investment in ecosystem partners, IBM is also expanding availability of its Cloud Engagement Fund (CEF), to all partner types, whether they build on, service, or re/sell IBM technology. CEF provides investment through significant technical resources and cloud credits for partners to help migrate customer workloads to hybrid cloud environments.

IBM's collaboration with Siemens Digital Industries Software is just one example of how the CEF is helping IBM partners scale.

Through this joint initiative, Siemens will apply IBM's open hybrid cloud approach, built on Red Hat OpenShift, to extend the deployment flexibility of MindSphere, the industrial IoT-asa-service solution from Siemens.

IBM also unveiled the world's first two nanometer chip which will enable faster, more efficient computing from the datacenter to the edge; Cloud Code Engine, a front-end platform that can help developers quickly deploy cloud-native applications without having to acquire new skills or configure complex code; Spectrum Fusion, a fully-containerised version of IBM's storage and data protection software designed to provide a streamlined way to discover data from across the enterprise; and IBM's alliance with Zscaler on Zero Trust where IBM Security Services combines the technology of Zscaler and the expertise of IBM to help clients adopt an end-to-end secure access service edge (SASE) approach, marrying authentication and permission to deliver security and privacy enhancements.

2ND QUARTER 2021 | SYNAPSE

51

NEWS

INVISIO AI SCOOPS 3RD PLACE at 2020

SAB Foundation Social Innovation & Disability Empowerment Awards

I

nvisio AI, a web-based software and patient application platform that uses artificial intelligence (AI) algorithms to detect breast cancer scooped third place at the 2020 SAB Foundation Social Innovation & Disability Empowerment Awards.

The platform, which was developed by MedSol AI Solutions won a R850 000 cash prize.The startup was founded in August 2020 by CEO and director Kathryn Malherbe (pictured ) who is a radiographer and mammographer with over 15 years clinical experience.

The South African startup has developed a patent pending world-first AI algorithm which uses deep machine learning (DML) to identify, segment and predict breast cancer types by means of ultrasound images and algorithms.

Malherbe told Synapse in late March that research and development work for Invisio AI had started at the end of 2019.

“There is a need to decentralise breast cancer diagnosis and treatment so that all women in South Africa can have access to on-site diagnosis, our solutions provides the diagnosis an Day 1 of clinic visit in the rural setting instead of the current standard of 6 months,” said Malherbe.

She explained that within two seconds of uploading an image onto the startup’s system, its unique patent-pending algorithm and ML model provides prediction of breast cancer subtypes without the need for on-site

52

SYNAPSE | 2ND QUARTER 2021

specialists of infrastructure for a histology.

“This means a patient in a rural setting can have a quicker turnaround time to surgical intervention, currently estimated at 4-6 months in SA,” she pointed out.

Malherbe said the platform was being piloted at three main breast radiology practices in South Africa, with its main target market being general practitioners (GPs), rural clinics, and district level hospitals.

“Our software aids in earlier detection of breast cancers, more cost effective detection of breast cancers as well as a reduction in mortality rates. We also promote breast education alongside Breast Cancer Support Pretoria NPO, who serves currently over 179 400 women on their digital platform across South Africa, Zimbabwe, Tanzania and Namibia,” she explained.

At the time of writing, Malherbe said the startup was in initial talks with companies based in Montpellier and Occitanie for its global outreach over the next few years.

Commenting on placing third at the 2020 SAB Foundation Social Innovation & Disability Empowerment Awards, Malherbe described winning the accolade as “greatest and most honourable achievement for our company to date”.

“To have a prestigious group such as SAB, not only see our innovation but our social impact is very encouraging for us.

The funding is opening so many doors for us, we can develop our current MVP in pilot testing to a fully fledged app as well as start our branding and marketing strategy for our proposed product launch in October 2021,” she added.

“We are revolutionising AI and DML to be an urban connectivity solution from grassroots level,” said Malherbe.

Check out this AI Expo Africa 2020 ONLINE talk by MedSol AI Solutions CEO Kathryn Malherbe on the Invisio AI platform: https://youtu.be/ yTaHlyhG60w

Entries open for ITU Digital World 2021 SME Awards The International Telecommunication Union (ITU) has opened applications for the ITU Digital World 2021 SME Awards, with submissions set to close on 3 September.

With a special focus on connectivity; smart cities (smart living); e-health, digital finance, and edtech, the awards not only recognise the most promising tech solutions with real-world social impact, but also celebrate the innovation and creativity of ICT-based applications and initiatives that are changing lives for the better around the world.

In addition, the awards provide an international platform to: • Highlight best practice • Share ideas on using tech for development • Network with peers, mentors and industry • Mobilise investment • Create new business opportunities for ICT solutions for social good.

The awards are open any startup or SME using tech creatively to solve socio-economic challenges anywhere in the world. Short-listed in front of an expert jury.

An expert jury will short-list applicants to go head-to-head in live online pitching based on social impact, innovative use of ICTs, business model, scalability, and environmental sustainability.

Winners across each category stand to benefit from: • UN recognition as an innovative ICT 4D solutions provider • International visibility before, during and after the event • Network with peers, industry experts, mentors and UN representatives • Participation in the SME Programme of capacity building and skills workshops, business matchmaking and networking events.

Apply for the ITU Digital World 2021 SME Awards here.

WIZZPASS WORKSPACE BOOKING

ADVERTORIAL

- Optimise & streamline your workspace management

M

any people have gotten used to working from home over the last year, but now that companies are opening their offices again, what does the future of our “workplace” life look like? Facility Managers, Health & Safety Managers, Property Managers and C-Level Executives are all faced with a whole new set of challenges to think about for a safe and effective return to the office.

How would you manage and optimise employee capacity per department, floor or per area? Which employees should be able to return to work? Which days of the week should certain employees be back at the office? How would an employee book a space or a desk?

A challenge or an opportunity?

These decisions can be difficult to make as it requires input from different stakeholders to find the ideal solution for each office space.

Many companies, however, have turned this thinking around and are seeing this as an opportunity to streamline their office space and better optimise their workplace. Now that employees are planning or have started returning to the office, this presents a great opportunity for companies to think about their own “workplace of the future” and in doing so, increase productivity and reduce costs in the longterm.

What is workspace booking ?

The concepts of “smart desking” or “hot desking” are not new, and have been around for many years. However, when these concepts were created before the Covid-19 pandemic, the first systems were built for a large and mobile workforce that would want the ability to book a space in different offices from day-to-day.

However, the challenges of Covid-19 and the new “hybrid model” of alternating workdays between the home and the office, has provided a resurgence for the need for workspace booking systems. Office-space

booking has quickly evolved and now allows you to allocate office-spaces according to Covid-19 social-distancing regulations, and present booking options to employees based on their seating preference, via real-time cloud-based software.

system. This coupled with insightful analytics on space usage allows administrators to gain the control and oversight that they need now and in the future.

Let employees choose

How do we monitor and deny access to an employee who has not made a booking, or who has not passed their Covid-19 screening? Should certain employees be allowed entry without having a booking?

These are important questions for each company or office manager to consider.

Workspace booking software such as WizzPass, not only allows for real time monitoring but also seamlessly integrates with access control systems. This allows for complete control of who is in the office at a certain time and day. Employee accessmechanisms at the turnstile or door (such as access card, facial recognition, etc) can be automatically blocked by WizzPass so that access to the turnstile or door is denied.

Going to the office might never be the same again, but that is a good thing By having workspace booking measures in place, it provides for more structure and control, but also fosters a democratic and transparent operating environment that allows employees to easily book and choose their own workspaces. At the same time, the company is setting itself up to optimise and streamline its workplace effectively, both for current and future needs.

WizzPass is a leading cloud-based workplace management system that allows you to effectively optimise your office space.

The system is flexible and easy-to-use and is used at all types of offices, including large corporate buildings. To find out more about the WizzPass System, please get in touch with us.

Giving employees the power to manage their days and times within the office not only allows for accountability and higher rates of productivity but also ensures that the highest safety protocols are being practised. The ability to administer seating plans effectively and quickly with cloud-based software is going to be instrumental in how businesses manage their offices – now and in the future.

These software systems should allow for immediate bookings and cancellations, thereby shifting the burden of administration away from office managers.

Optimisation and collaboration Administrators should have oversight on which employees intend to be in the office at a particular time and should be able to allocate certain employees to book only in certain areas. This allows for effective planning and increased collaboration on projects between different departments and people.

Managers would no longer need to worry about paper trails, and can now manage capacity, health-screening, desk-cleaning frequency and collaboration-planning on one

Flexible settings, but with secure controls

2ND QUARTER 2021 | SYNAPSE

53

2nd QUARTER 2021 | ISSUE 12

SYNAPSE Africa’s 4IR Trade & Innovation Magazine

NVIDIA INCEPTION: 3 African Startups Accepted into the Programme

TUNBERT: 1st AI-based

Tunisian Dialect System

3 AFRICAN STARTUPS using AI, Data Science for Financial Inclusion

AFDB BACKS AI-BASED

National Consumer Management Systems

HOW THE PANDEMIC Gave Birth to SA’s latest 4IR SaaS platform

LACUNA FUNDS DATASETS for Low Resource African Languages

SYNAPSE

2nd QUARTER 2021 | ISSUE 12

SYNAPSE Africa’s 4IR Trade & Innovation Magazine

Africa’s 4IR Trade & Innovation Magazine

NVIDIA INCEPTION: 3 African Startups Accepted into the Programme

TUNBERT: 1st AI-based Tunisian Dialect System

REACH AFRICA'S LARGEST ARTIFICIAL INTELLIGENCE & 4IR COMMUNITY WITH SYNAPSE MAGAZINE

3 AFRICAN STARTUPS using AI, Data Science for Financial Inclusion

AFDB BACKS AI-BASED

National Consumer Management Systems

HOW THE PANDEMIC

Published Quarterly

Gave Birth to SA’s latest 4IR SaaS platform

LACUNA FUNDS DATASETS for Low Resource African Languages

Official Publication of AI Expo Africa

RATE CARD RATE*

AD SIZE Company Listing (logo, company description & hyperlink) 1/4 Page (includes 1/4 page editorial) 1/2 Page (includes 1/2 page editorial) Full Page (includes full-page editorial feature) Double Page Spread

R2500 R3500 R4500 R6500 R10,000

* All rates exclude VAT & agency commission. Rates are based on casual advertising.

Discounted rates are available for longer ad / editorial runs

READERSHIP / SOCIAL MEDIA REACH Synapse Magazine is Africa’s first and only business quarterly publication covering developments across the continent in Artificial Intelligence (AI), Data Science, Robotic Process Automation (RPA) and Fourth Industrial Revolution (4IR) smart technologies.

Synapse offers industry executives, practitioners, investors and researchers relevant news, in-depth analysis, and thought leadership articles on trends around 4IR innovation and digital transformation in industries that include banking, retail, manufacturing, healthcare, mining, agriculture, education, and government, among others.

Over the years the magazine has established a significant following across Africa as well as globally, with readers from as far afield as the North America, South America, Europe and Asia. This makes Synapse a great marketing platform for startups and established tech companies to reach a broader community of buyers, investors and partners.

Readers around the world

With its insights, interviews and case studies, the magazine aims to be a voice for African 4IR practitioners, researchers, innovators, thought leaders, and the wider African AI community.

Since its launch in 2018, Synapse has amassed a combined readership of 31,300 across the Issuu platform (on which it is published), the AI Media Group’s email database, the AI Expo Africa Community Group on LinkedIn and the AI Media Group’s social media channels where the magazine is distributed. It also links to AI TV, Africa’s only dedicated YouTube streaming channel focused on 4IR business users and trade.

BOOK YOUR AD / EDITORIAL FEATURE

Advertising and artwork to be supplied as a high resolution PRESS-ready PDF of at least 300dpi.

Art and editorial features to be submitted to: daniel.mpala@aiexpoafrica.com