In the the biggest webinar of the new yearthree credit experts went live Thursday to the LendIt audience to talk about the future of AI analytics in credit.
The casting was particularly requested by the public of LendIt. After a surprising year of fintech startups and new products, it’s refreshing to tie new tech to credit and underwriting.
Srikanth Geedipalli, the SVP of Global analytics and AI at Experian, joined Michele Raneri, Data Product and Risk within the startup BNPL Opy, and Murli Buluswar, Head of Analytics US consumer banks at Citi. Host Peter Renton led the trio to explore the next generation of credit.
Geedipalli began by introducing himself and the panelists.
“Welcome to all of our viewers today. Hope this has been a great start to your new year,” Geedipalli said. “About my role: I am the Product Manager for Analytics Products for the experienced global team. My mission here is to “produce” the analytics, which is to take a long enough analytics project and turn it into a fast, better, faster and seamless delivery of the solution to our customers. »
How Experian handles chargebacks and builds customer loyalty
Geedipalli, who worked at McKinsey, Capital One and BMO for more than 16 years before leading the Expiry AI product teams reported using predictive technology and machine learning and artificial technology (MLAI) to help build customer loyalty.
Renton asked how the three handle chargebacks before and after and how MLAI helps with delinquent accounts, but as Geedipalli explained, the last two years have seen a strong growth market with no issues with delinquencies.
Instead, the problems of a bull market stem from keeping customers when there are so many other options.
“There are a lot of opportunities on account management, how to keep them engaged, how do you make sure that when they get another card offer, they don’t walk away?”
Geedipalli also said the answer applied to the original question: “So think ahead, rather than react to it.”
Getting in front of customers who are on the verge of leaving, about to miss a payment, and offer them products and solutions that keep them on track and, most importantly, keep them in-house.
“So there’s a bunch of opportunities just closing and talking specifically about that, okay, so customers,” Geedipalli said.
“Users who actually use the product and are close to a charge or default, these are both opportunities for lenders to get ahead of the problem, offering solutions before they do not find themselves in a state of delinquency.”
Machine learning is required for credit models
Raneri has 30 years of experience leading consumer and business credit market analysis at Citi and Experian and recently joined Opy. In this young BNPL fintech start-up, she leads data and risk strategies.
She said she couldn’t comment on the bigger partnerships coming out the following year, her silence indicating hope that Opy will form a partnership in 2022, but she had a lot to say about the analysis of the IA in BNPL.
Raneri focuses on creating BNPL connections with clients’ current accounts – Demand Deposit Accounts (DDA).
There are so many Americans with non-traditional finances, and building credit models for them requires data innovation. A third of Americans work in the gig economy and their earnings are not regular, Raneri said.
“I have a lot of experience with this DDA data and building flow analysis. When you think about what your DDA looks like, what your checking account might look like, a lot of us have regular income, salaries and we get paid very regularly,” Raneri said. “But what we’ve found is that there are so many people who say in the gig economy or get paid irregularly.”
Opy and anyone trying to build products need to develop models to calculate the new non-traditional market, and Raneri said the complexity required machine learning.
“These attributes really needed this sophistication of machine learning in order to have so much data and still be able to develop attributes on influx and influx types and to be able to get more granularity so you can build it into models,” Raneri mentioned.
“It was absolutely fascinating. So if anyone is interested in doing this kind of cash flow analysis, machine learning would be the only way to do it because of the many different permutations that people deviate from.
At the top means on tiptoe: reacting to change
In 2020, call centers around the world shut down when customers needed support the most – that’s when Buluswar and the team launched a predictive technology solution.
“How do you avoid a situation or mitigate the situation with customers having to wait for hours and hours,” Buluswar said. “What we needed to do, based on customers’ digital behaviors, is predict who would call, why they would call, and find ways to use advanced analytics to intercept their need to call.”
Templates have saved time in support, but the fundamental goal is to apply the idea of time saving to everything: consumers need a suite of features that quickly anticipates their needs. For example, the normal six-month lead time for engineering new applications needs to be reduced, Buluswar said.
“Our challenge has really been how to use increased speed and cycle time to generate insights and reduce the obvious friction that exists,” Buluswar said. “This cycle time must fundamentally decrease. It reduces this friction. This allows us to build a library of features that we can draw from. It’s the keystone that transforms the speed and sophistication with which my team delivers information. »
Predictions for the new year
Renton asked the three credit experts for their views on the most important products coming in the new year in a quick final section at the end of the conference.
Srikanth: “Cash flow underwriting, using more data, using current account data and transaction data in your underwriting models.”
Raneri: “Buy now, pay later” and as a key regulatory theme: Continued focus on credit data privacy.
Buluswar: “Connecting the lender to the borrower much more effectively and coping with the impact of rising interest rates.
