7 minute read
Beyond ISO 20022
Beyond 20022
The enriched data that will flow from compulsory adoption of the new messaging standard will be a ‘protein boost’ for AI in the field of transaction analytics and forecasting. Victoria Harverson from SmartStream believes banks need to seize the opportunity
SmartStream’s trademarked approach to solving operational processing challenges – known as Transaction Lifecycle Management (TLM) – had something of a head start on what’s been described by one leading bank as ‘a watershed moment’ for the payments industry.
It’s been working in the ISO 20022 environment since the payment messaging standard first emerged in 2004. However, as compliance with ISO 20022 becomes mandatory over the next few years for any institution transacting through EURO1/ STEP1 and TARGET in Europe, processing SWIFT messages, or using Singapore’s MEPS+, Hong Kong’s CHATS and Australia’s RBA-RITS clearing, the heat is on to bring every one of its clients up to speed.
The logic and aspiration behind ISO 20022 is irrefutable: it’s designed to enhance the speed and efficiency of global payments by enforcing universal uniformity, which will demand a new granularity of information around every transparency-boosting messaging systems like SWIFT gpi have left off, by ensuring not just increased payments visibility, but also real-time speed and glitchless delivery.
SmartStream’s considerable technical fire power helps organisations manage their data for in-time compliance, and ensure exceptions are identified early to minimise disruption. Its services include data transformation, regulatory alignment, innovation labs, managed services, Cloud environments and solution delivery.
Global head of business development, Victoria Harverson, believes organisations must embrace technologies like AI, to make the most ISO 20022. Too many, she says, overlook AI’s power to transform their behind-the-scenes activity. Banks need to decide how they will manage their – often outmoded – core architecture as they adapt, to avoid their back-ends limiting their capability for frontend innovation.
To help them, in March, SmartStream announced it had re-engineered its TLM Aurora software-as-a-service solution to help banks upgrade their core systems to comply with ISO 20022’s more complex data requirements. TLM Aurora Universal Data Control allows for basic-to-complex reconciliations data matching by enabling AI, machine learning and Cloud-native technology.
financial message, spanning not just cross-border retail payments but also trade, securities and foreign exchange.
Importantly, for a company that has made a strategic shift to harnessing artificial intelligence, ISO 20022 is a protein-boost for AI. The rich information attached to every transaction will enhance analytics, improve reconciliation rates and – the end game – improve the experience for the customer. But migrating to the new standard is not simple if your payments engine is built on legacy infrastructure.
To support clients, SmartStream expanded its TLM Corporate Actions solution in 2019, to include ISO 20022 message processing and interoperability support. Its new iteration of its SmartStream AIR Cloud-native data validation software solution, and the new TLM Aurora Universal Data Control platform, both powered by machine learning and AI, are bringing the latest technologies to help companies migrate to and maximise the opportunities that the new standard presents.
Under ISO 20022, every character within a financial message must be 100 per cent correct and aligned with the specifications, and the format is validated at several points along the communication chain from the sending and receiving ends. A single missing colon could result in a multi-million currency transfer being rejected or delayed for days.
But, once compulsory, ISO 20022 could revolutionise the increasingly-borderless payments marketplace, picking up where
Speaking at the time, Roland Brandli, strategic product manager for SmartStream, was quoted as saying: “The discussions we are having with banks right now revolve around ISO 20022… where financial institutions are having to upgrade legacy systems for data remediation and improved channel and back-office processing. We decided to [develop] a new environment to meet today’s business needs. [It] goes over and beyond basic reconciliations… and is based on open architecture and adaptable to all banking requirements, without the need to re-engineer operational processes.”
In April, SmartStream also launched V3 of its proprietary AIR software-as-a-service (SaaS) platform, designed to serve organisations worldwide that are using Microsoft’s Azure platform for data management and infrastructure improvements, as well as AI solutions access. Also hosted on Amazon Web Services (AWS), the move to Azure offers users more flexibility to use regional data centres for their localised compliance needs – as well as data lake support for improved upload, handling and manipulation, and customised reports offering deep reconciliations insights.
Harverson says the company is well aware of the opportunity payments represents: “At the moment, most of our experience lies in capital markets and use-cases within key financial services, although we are moving much more into payments, and the Payment Card Industry Data Security Standard is key for us.”
The company achieved PCI DSS accreditation in 2020.
Harverson advocates a balance between core infrastructure improvements and innovation. “Banks are constantly striving to compete in an age of disruption. There’s lots of demand for incredible front-end services, focussed on customer experience, and a dedicated function and budget for tech innovation activity is key because the client experience is their differentiation and biggest profit contributor,” she says.
“However, if they over-focus on those things and don’t invest in their underlying systems, including end-of-lifing old components and rebuilding their architecture so that it’s flexible, scalable and Cloud-ready, they risk being disrupted by the competition. Our customers, using our AI solution in their core banking architecture, can easily inspect data between front, middle, and back-office systems. They can also better leverage their transaction data, then run business intelligence to do things like forecasting client spending. We are also running initiatives with banks’ technology stakeholders, to help them speed up their transformation projects, using AI and other tools for doing things like quality assurance tasks that would usually take hours, in minutes, with a fraction of the resource.”
But AI must be used wisely to bridge between old and new. “Seeing the value in AI, doesn’t always mean an organisation has
Smashing barriers:
AI should be a key enabler for banks
transaction data, in different formats, in minutes, rather than days. We focus on doing more with less.” But adopting it, for banks, can be tricky. Harverson adds: “Our Cloud-native SaaS product is easy to use; with the right security, standards and certifications to allow clients to run any data on it, easily. “But reduction of team resources and changes to governance can be real issues. A bank might have used a multi-step data management approval model for years and it’s difficult to move from that to an automated, faster solution. “AIR offers audit logs and capabilities, so we can export details of how processes have been performed, or the client can see it all on the user interface. Fintechs like us, and banks, are trying to do things to make life easier, but traditional models and internal rules that have been in place for a while can make enabling change difficult.” SmartStream sees itself very much as part of a cavalry of fintechs. “We use AI to reconcile data rapidly and effectively, and do that really well. But there are other fintechs that do data visualisation, for example, equally well. In the front We can integrate to those an efficient implementation end, people ‘get’ AI … But and allow our customers to run deeper analytics and business intelligence. strategy for it,” says in the back end, “You have to focus on Harverson. “Legacy IT infrastructure and systems often don’t share data or it can control risks, increase your strengths and let others focus on theirs. It’s important organisations interoperate, and there are productivity and think about the strategy for still divided opinions around using AI. However, if banks don’t do this, others will. In efficiency, and lower costs such ecosystem partners.” Therein lies the potential for further the front end, people ‘get’ iterations of AIR. “We’re AI and we talk about ultra-personalised talking about integrating our AI with experience, like the Amazon analogy. But in business intelligence and management the back end, AI can control risks, increase information service tools that allow efficiency and lower costs – doing things like organisations to visualise reconciled AI fraud prevention and identity verification.” data more meaningfully. We’ve built the
One of AI’s greatest benefits is efficient product and are selecting providers to data management, making it indispensable work with,” says Harverson. So, what next? for ISO 20022 readiness. “Our innovation “I’m looking forward to more labs focus on using AI to compare opportunities within open banking; API high volumes of data and find platforms and ecosystems of applications discrepancies instantly. There’s no need and toolkits for financial institutions to to build data reconciliation; AI does that. access and integrate fintech products from We use supervised, unsupervised and different companies in one place, with no observational learning techniques, linking code needed,” says Harverson. system data together to make sure it “We continue to listen to customers displays in all the systems accurately. and build our internal agility, so we can And we can reconcile vast volumes of deliver future-proof solutions.”