TRANSCRIPT
EPISODE 26: Supriya Gupta
Jim Freeze Hi! And welcome. I’m Jim Freeze, and this is The ConversAItion, a podcast airing viewpoints on the impact of artificial intelligence on business and society.
[UPBEAT MUSIC]
Today, I’m joined by Supriya Gupta, Senior Product Director and Head of Recommendations at Credit Karma, a personal finance company that provides educational resources, tools, and personalized recommendations for credit and loan management.
Supriya has played a critical role in scaling the development and adoption of AI technologies throughout Credit Karma’s business. We discuss everything from the company’s latest machine learning innovations, to its vision of creating a mobile, personal financial assistant that supports consumers across the whole financial landscape.
Welcome, Supriya! We’re thrilled to have you on the show.
Supriya Gupta Thanks for having me.
Jim Freeze We’re very excited to have you. Let’s start out with a little bit about your background. You have extensive technical and business experience at both startups and global enterprises. What drew you to FinTech and Credit Karma in particular?
Supriya Gupta Yeah, I’ve spent a number of years working at various companies, most of which have a strong AI component to them. The attraction to Credit Karma for me, was really the chance to apply my years of expertise to this wonderful mission of Credit Karma to champion financial progress and also to be part of this mission-driven company and culture, which has this true intention and the product experiences that reflect that intention, that really enable and empower people in a financial way. There’re just endless opportunities, specifically in finance, in terms of problems that need to be solved in the society at large. Credit Karma is just so well positioned to solve these problems, especially using AI, so, this was the attraction for me with Credit Karma and FinTech.
Specifically, AI can have a really positive impact in this space in terms of leveling the playing field for consumers by fueling choice and certainty and all sorts of efficiencies. This vision that our CEO, Ken Lin, talks about around automating finance was just this really attractive opportunity to be able to leverage the power of AI and data for good. That’s what brought me to Credit Karma and this amazing opportunity in FinTech.
Jim Freeze That’s great to hear, specifically a mission-driven company that’s really helping with financial literacy, which I think is a growing issue for many folks, so I completely get the excitement about what you’re doing for good. Furthermore, a little bit diving into Credit Karma, from what I understand, Credit Karma took off in the height of the financial crisis back in 2008 to help consumers manage their credit scores. Since then, it’s evolved into a more holistic FinTech company that offers support across credit, loan, mortgage management and I think as you highlighted, AI is playing a significant role in that evolution. Can you walk us through the shift and the growing impact of AI?
Supriya Gupta Yeah, absolutely. Just to start from the beginning here, the early days of Credit Karma, when our founders started this company, this was at a time, 2008 again, when consumers didn’t have access to the information banks had on them. Their credit scores, credit reports, were not easily accessible and the founders really knew, deep down, that this was just fundamentally the wrong way to build this industry. That this transparency was really important, and it was really clear to them early on that credit scores were really a means to an end for most people and the prosperity in their lives. Those looking to improve their credit scores… It was not just the score itself, it was in pursuit of something bigger, like getting a credit card to be able to purchase more things or even getting their first car or home. We started to level the playing field here with free credit scores and—
Jim Freeze Which by the way, I have to say, I took advantage of at the time. It was great because to your point, there wasn’t transparency and there was this mystery around credit score and if you wanted to know you had to pay for it, so I think what you guys did was fantastic. I’m sorry to interrupt your flow there, but I remember that and taking advantage of it.
Supriya Gupta Yeah, absolutely. I still remember the jingle from their early commercials back in the day, well before I actually joined the company, about free is free and I—
Jim Freeze Sing it for us, go ahead and sing it for us.
Supriya Gupta Maybe some other day, after a beer or two at happy hour, but yeah, no. It was really amazing. I still remember you had to pay 20 bucks or something to get your credit report, it was a whole thing. It’s kind of amazing that that industry was able to do that for so long. So just to continue on there…yeah, so that was the humble beginning, was flipping the script and just really democratizing access to what people fundamentally should have had access to in the first place.
But where I was going with this was, the vision for this was always much greater. It was free credit scores at the beginning, but it was really that as a conduit to be able to give consumers a fair chance to make financial progress, really providing that transparency, certainty, and simplicity around finances to achieve that. As the business grew on top of this great business model and value brought to our consumers, technology actually also got more advanced as well. Specifically when it comes to AI, the cloud just evolved tremendously since 2008 and technologies became more accessible through the cloud and we were able to leverage that technology in ways to build out AI in ways that 10 years ago, 20 years ago, would have just been exclusive to juggernauts like Google or Amazon.
We started our recommendations team about seven years ago and our objective starting out was to first stand up offers delivery, and then it evolved over time to be this dynamic, personalized experience that could recommend anything, and all this fueled by data and machine learning. Like I mentioned, we started off with offers, recommendations, and certainty. For Credit Karma, we were the first company to really do these approval odds at scale and we grew from there and evolved the AI.
Just a quick definition moment for those listening, that may not be as familiar with the FinTech terms, so when I say offers, what I’m referring to is offers from our financial partners for financial products, like credit cards, personal loans, et cetera. That’s what offers means and when I say personal loans, this is a product we have on Credit Karma and it actually uses AI to predict the approval odds a member has for specific offers on Credit Karma, based on what we know about your credit report. This allows us to rank offers and highlight offers that the consumer has higher probability for qualifying. This is really, really important and really—
Jim Freeze Very.
Supriya Gupta … powerful, because if you apply for products you don’t qualify for, then you take a hit on your credit score, but you still need the product. Now you’re disadvantaged when you apply again, and so this was one of the really big innovations we had early on in AI.
Jim Freeze Does what you’re talking about, I think you launched a kind of an AI-powered feature last year called Stories, which expanded the app’s capabilities, particularly in your domain of recommendations. What was the inspiration behind that? If you could talk a little bit more, specifically about the role of AI in that, that would be fantastic.
Supriya Gupta Yeah, absolutely, so offers and certainty were our beginnings and like I’d mentioned, we evolved the platform to cover every single, aspirationally, every single pixel in the app. Stories specifically, was our foray into that expansion and this is effectively a feed-like feature that you see when you log into the app, on the landing screen. What this does is, it serves up the most relevant information, recommendations, tools, offers, all sorts of things to help them achieve financial progress and these stories are tailored for their individual situations, most of which are actually personalized with information that is actually specifically relevant to that member.
It was a really, really powerful addition to the app, because it not only surfaced recommendations, but it surfaced them in a way that was actually contextualized to the member. Where they are, what they need, and what their situation is, and it allowed us to actually create a much richer discovery experience to find things within Credit Karma, because as a company we actually offer a lot of different tools, products, services, different types of financial products as well, but different people have different needs. You could take a static approach to navigation, which is what we primarily relied on before, but Stories allowed us to drive personalized navigation, to servicing different parts and experiences of our app that are actually much more relevant to where members are and what they need.
I’ll just give a quick example of this, so back about a year ago or so, when we launched Stories, we surfaced Stories about mortgage refinance and through that effort, we were actually able to drastically raise awareness of this product space in general. People who actually got served up mortgage refi offers by us grew by 4x, and the people who actually took them grew by 6x, which is tremendous.
Jim Freeze That is. I mean, it speaks to, I think the value and the relevance of what you’re delivering, right?
Supriya Gupta Yeah, absolutely.
Jim Freeze Oh, that’s fantastic. I lived through 2008 and I thought, “Boy, that’s you know…” The circumstances behind that and the impact it had on the financial market and people’s concern about personal finance. I thought, “I’ll never see anything like that again.” Then fast forward to 2020 and a pandemic, and I’m thinking a year ago in late March and April and the market was tanking and I got to believe it’s had a huge impact on demand for financial support. Is that true and how did that shift affect what Credit Karma was doing?
Supriya Gupta Yeah, so it’s definitely true. Last year, especially right at the downturn at the beginning of the pandemic, we saw nearly 40 million members visiting Credit Karma each month. It’s a really, really high engagement. They were looking for relief from a very unique set of financial challenges that were brought on by the pandemic, specifically.
Jim Freeze Yes.
Supriya Gupta It actually was very interesting. It was a very interesting AI problem in fact, to transform and personalize the app to surface the right things, in the context of what was happening in the environment. There were a couple of things that I can mention in this category. One was around just the underwriting and certainty related recommendations that we do that reflect that underwriting. Those rapid changes, that required us to actually evolve and rapidly update our AI predictions and surface accurately and currently what sorts of offers are actually available to our members, at the best possible rates. Especially considering how underwriting was tightening across the ecosystem for many of these banks as people were losing jobs and things of that nature, so that was really turbulent, but really, really important work that we took very seriously at Credit Karma to surface that.
If we take a quick example back to the refi thing that I’d mentioned earlier, mortgage refi, so REX has actually helped surface this super powerful opportunity to save money. When many Americans were strapped for cash, being able to go do a refi on your mortgage meant saving hundreds of dollars for many people, which was a difference between making other bills and things like that, especially as rates were going down to record lows last year. That was part of what played into that and our ability to shift and provide that opportunity.
That was one, and then the other one was a really interesting consumer experience we actually built out, called the relief roadmap. As banks, lenders and government agencies started to respond, we saw this influx of relief information and programs flooding the web, and it was effectively almost impossible, it was super complex to try to navigate, literally hundreds of new programs that were popping up to help people. We knew that that was going to be top of mind for many of our members to try to find the most relevant and applicable information and resources for their individual situation. For exactly this reason, we were able to use our internal content platform technology to actually stand up an experience within weeks, that allowed people to find relevant and applicable information and resources to their specific situation, depending on where they lived and their circumstances, exactly what programs they can go to apply for and get some information on how to do that.
We used AI to not only surface these recommendations for this type of relief content, but we also used AI to drive notifications to these members and help them track progress and encourage them to continue to make progress by leveraging all of these resources. This was also like a tremendously successful effort. We had nearly 20 million members turn to this specific product for navigating their finances during the pandemic.
Jim Freeze Boy, talk about a real demonstrable example of the value of AI and how it can help people in circumstances that we’ve all experienced over the course of the past year. Thank you for sharing that. That’s great. Now, I know back in December, Credit Karma was officially acquired by Intuit. I read an article in Fortune that discussed the two companies’ shared vision for creating a mobile personal financial assistant to support consumers across the entire financial spectrum. Can you speak a little bit more about that vision?
Supriya Gupta Yeah. Well, our CEO Ken Lin is probably the best to speak about this shared vision, but it’s certainly grounded on the idea that we’re better together. We’re better positioned to reach our vision of automating finance through a personalized product that will ultimately help large swaths of our American society save a lot of money and be prosperous in their financial lives. With that said though, we actually do operate independently from Intuit, but there are a lot of data synergies we plan to tap into to accelerate this vision. Specifically what I mean by this is, building on our deep understanding of liabilities. If you think about a balance sheet, right, there’s assets and liabilities. We have a deep understanding of liabilities because we have credit bureau data, but we can now have a deeper understanding of the other side of the balance sheet, our members’ assets like their income, and other things that Intuit has a better sense of.
At a high level, these data synergies will help us understand the holistic picture of the finances of each of our members and then use that holistic picture to optimize, and hopefully provide cheaper and better financial products on the whole, while simultaneously doing even better at helping them understand which products they qualify for and why, and how to improve their financial profiles as well. So, as a recommendations team specifically, we plan to leverage the Intuit data and features to improve our recommendations, as well as the member experience in this way. Then, also explore even more types of financial recommendations that we can’t make today, because we don’t have the insights that we now do from Intuit. Specifically, to truly understand how to automate finances in that tailored way that really optimizes for each member’s financial outcome, we have to have that full picture of the financial situation and journey and so this is how I see Intuit helping us get there, as Credit Karma.
Jim Freeze Yeah. It sounds like they can really help you execute on…I wrote this down, I love this expression, from free credit scores to financial progress. I think that’s a great explanation of the journey that you’ve been on for the last dozen years or so. To kind of wrap things up, I’d love to get your take on what’s next. How do you see the role of AI evolving into the financial services industry over the next 5 to 10 years?
Supriya Gupta Yeah, so AI is…it’s interesting, you’re seeing it more and more making its way into traditional financial services, which I find interesting. When you think decades ago, banks used to do their underwriting based on very, very basic criteria. Now, most banks are moving towards AI driven credit underwriting, and we see this increasing more and more over time to smaller and smaller lenders as it becomes more accessible and easier to do. With that, I’ve seeing better assessment of borrower risk using data-backed assessments, so machine learning combined with alternate data, very sophisticated fraud detection as well; that enhances your risk assessments. I’m also seeing some lenders go deeper into the exploration of very personalized offers and incentives, definitely seeing that trend. I see that happening more and more, in much more granular ways. Those are some of the trends I’m observing.
One of my hopes is also, and I imagine this will happen in the coming 5, 10 years as you had said, is AI-powered fair lending assessments, or just generally fairness applied to these processes where I think traditional finance can continue to evolve and be better. I also see this as an interesting challenge as lenders also dive into more personalization on the financial side. That’s a little bit about the sort of evolution of traditional finance into the FinTech world. Then, maybe more of the FinTech first or folks in companies like Credit Karma and alike, I could also see the tailored recommendations, financial recommendations, like the ones we have at Credit Karma, continuing to grow across the financial ecosystem because, as we’ve learned, it’s really democratizing the access to these financial services in this ecosystem; in this way it truly levels the playing field and helps society as a whole prosper, and these types of tailored recommendations are going to be what allows for that democratization to happen.
Jim Freeze Hear, hear. Hear, hear. I totally agree with you. That’s great.
Supriya Gupta Yeah, and then just the last point here on Credit Karma specifically, I think my hope is that in 5 to 10 years, we see that vision that Ken Lin, and the leadership talk about a lot, for autonomous finance and putting consumer finance into autopilot. Eventually, most of our important financial decisions could be automated, or at least the complex decision-making and the demystification of finances, becoming a real thing and really making it much more accessible for consumers. That’s where I see the next 5 to 10 years for AI in general, and using AI and data for good in this way.
Jim Freeze That’s fantastic. It’s been fantastic having you on The ConversAItion, talking about the growing role of AI and Credit Karma’s services more broadly and personal finance. We really, really appreciate it. Thanks for joining us and our listeners, I know, will love this episode.
Supriya Gupta Thank you. Thanks for having me. This was really great.
Jim Freeze On the next episode of The ConversAItion, we’ll speak with Alison Darcy, Founder and President of Woebot Health, about combining AI and digital therapeutics to automate therapy, and make it radically accessible to ensure patients receive the right intervention at the right time.
This episode of The ConversAItion podcast was produced by Interactions, a Boston-area conversational AI company. I’m Jim Freeze, signing off, and we’ll see you next time.
[UPBEAT MUSIC]
###
