SmartStream
AI and ML spell innovation
Artifi cial intelligence and machine learning have become much more than simple buzzwords for fi nancial services. Rather, they have become integral to the way banks analyse data and derive value from it. Jethro MacDonald, product manager for AI and machine learning at SmartStream, tells Jim Banks how the application of AI and ML continue to increase the functionality of key post-trade solutions.
ince the foundation of SmartStream’s Innovation Lab in 2016, artificial intelligence (AI) and machine learning (ML) have been at the heart of its efforts to develop more sophisticated functionality for the company’s solutions. The lab exists to proactively investigate and evaluate emerging and disruptive technologies – and both AI and ML are central to that mission.
S
Jethro MacDonald, product manager for AI and ML at the Innovation Lab, oversees the application of both technologies across SmartStream’s product portfolio and is continuously investigating how they can be applied to both current challenges and core issues that will emerge in the years ahead. That work has led to major steps forward in the company’s current offerings to its clients.
“There are three main areas that we are looking at,” MacDonald explains. “The first is the onboarding process for our reconciliation solutions to speed up that process. Previously, that process was manual and time-consuming, but we have developed Lightning to correlate data from different data sets, define how they should be linked and how we can identify matching rules.”
Lightning strikes
Lightning is the latest adaptation of SmartStream Air, the company’s flagship reconciliation platform, which grabbed the attention of the banking industry because of the sheer speed at which it can match data sets of all kinds – both structured and unstructured. SmartStream Air raised the bar on the reconciliation process, enabling banks to match data sets in seconds rather than days or weeks. Yet, the process of creating sets of rules for matching data remained a sticking point, requiring both time and manual effort, as well as foresight and experience, to anticipate where problems might arise. AI helps Lightning circumvent that process. “Lightning goes through all the data, links it together and creates the match rules automatically in a very controlled manner,” says MacDonald. “People worry
18
about AI being in a black box, so we show at each step why we think the data should be linked together in a certain way. Clients can keep or change those rules. “Lightning compresses the time and effort needed to create rules and controls,” he continues. “That is important because our clients want to build controls much quicker. This is a logical solution for AI, and it is actively used by many of our clients.” MacDonald is in charge of SmartStream Air and of innovation processes at SmartStream more broadly, and a key focus of his work is the integration of AIR into all of the company’s core systems – benefitting both new clients and customers that are signed up to its existing solutions. A key development of Smartstream Air is the incorporation of Affinity, which tackles the second of MacDonald’s core challenges. “That is the minimisation of the manual work involved with data matching,” he explains. “Once the reconciliation system is built and the onboarding is done, then the effort for manual matching needs to be reduced. Affinity is a learning AI. It learns from a user’s behaviour, proposes suggested matches for unmatched data based on what the user has previously done. “Once again, everything is shown to the user who sees the substrings that go together and why. The user can confirm those matches and the solution gives a level of security, transparency and visibility. This is not just AI in a black box.”
Over the past two years, the turnover of staff in the industry has been high, not least because of the effects of the Covid-19 pandemic. And when key people leave an organisation, they often take their knowledge with them. By learning from user behaviour, Affinity retains that knowledge within the system itself. Ultimately, people can leave and the system retains their knowledge, and the training of new people is less time-consuming.
Taking exception to errors Having accelerated the onboarding of data, dramatically boosted the speed of reconciliation and
Future Banking /
www.nsbanking.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53
