The hunt is on for the best use case for AI. But instead of looking for a one-size-fits-all AI application, firms should first focus on an incremental approach, taking small steps that add up to massive improvements, says Timothée Raymond, Head of Innovation and Technology at Linedata, a Paris-based company that provides technology solutions and services to the global asset management industry.
What impact will AI have on financial services?
It will disrupt and change the industry at its core. But instead of talking about what it replaces, we should focus on how it will enhance the way we work in the future. The greatest value to the industry right now is not transformative, but the combination of many smaller changes. It’s what I call the ‘power of five seconds’. Optimising every single interaction and making small improvements within a system adds up to a big difference.
Can you provide examples of these incremental improvements in financial services?
Existing compliance processes typically generate a number of false positives. With a dynamic approach, AI can learn from the past and eliminate some of these false positives, even without defining its actions as a rule. Another example is speeding up internal processes. Some problems can be solved by an analyst, but sometimes you need an experienced senior accountant. An AI that can route questionable issues to the most appropriate team member is hugely beneficial. Of course, humans could do this in seconds or minutes. But having an AI that continuously improves the microscopic efficiency of the operation will ultimately change the way banks operate and organise themselves. AI will not solve every single problem, but it is part of a bigger picture.
“AI will disrupt and change the industry at its core.”
How should BFSI organisations take their AI journey to the next level? Alone or in partnership with specialists?
How will software testing change as AI advances?
Testing has changed very little over the last 20 years. But traditional testing methods are not suited to the complexity of today’s systems. Adding more people is not the answer and has never solved a problem in the long run. If we want to improve operations, we need to rethink the way we test. Implementing an AI-based self-learning system is an excellent option. In rule-based systems – be it compliance or testing – we work with predictions of what can go wrong. The problem is that, by definition, most of the time things don’t go wrong the way you thought they would. A system’s ability to be more creative or explore ideas that might seem irrational to humans adds significant value. The ability to be dynamic in such a challenging environment is something where AI outperforms humans.
How important is AI for Linedata?
It’s very important and one of the strategic pillars of our five–year plan. We are looking at AI from two perspectives: One is to help our customers deal with technological advances and potential disruptions. And for that, we strongly believe in the five-second improvement, which is why we are committed to building AI into every single product we have within the next 12 months. The other aspect is inward looking: how we are changing as a company and how we can improve as a solution provider. Our developers are using copilot systems and other technologies to deal with their code. We are also training people with AI and changing the way we look at technical management of people or testing of applications.
“The success of AI depends on greater openness and data sharing. Change is inevitable and the industry cannot afford to be too defensive.”
This interview was taken from our recent report How AI is shaping the future of financial services.
