Prem Natarajan of Capital One, interviewed on stage at HumanX, explained how generative AI is transforming the lives of their bank’s customers in terms of access to finance.
Capital One is one of the largest and most renowned American commercial banks, with over 7,000 ATMs and 750 branches.
Surely this is still an early stage from which everyone is learning, explained Natarajan, but the first application of AI agents that Capital One has used is applied to customer support, called “chat concierge”, which seems to have improved customer responses by 93%.
Capital One is also using AI agents in sales management, increasing contract closures by 55%. The improvement percentages are given in comparison with previously used solutions.
Capital One has a sandbox that tests AI solutions, and many of these are obviously tested and discarded before reaching the public.
AI to Prevent Fraud
AI is also helping to prevent fraud, for instance, when a credit card is swiped on a reader.
“A significant step forward in combating scams in the financial sector is when the transition was made to using machine learning by learning from past data and updates arrive within a couple of days instead of months. And it is a highly scalable mechanism.”
AI and Regulation
There are many regulatory agencies, and each bank constantly meets with them to discuss what is being done and also future plans. Developing a smooth and transparent communication with the regulator is important, but since machine learning is something Capital One has been working on for years, they have managed to anticipate the agencies’ requests.
“We are able to make steady progress thanks to our ability to communicate with the regulator and anticipate their requests”, explains Natarajan.
