PitchIt Podcast #9: Justin Wickett of Informed.IQ

In episode nine of PitchIt: the fintech startups podcast we talk with Informed.IQ Co-Founder & CEO Justin Wickett.

Artificial intelligence or AI has quickly become one of the most popular buzzwords in all of fintech. Companies constantly highlight their use of AI, but there are only a few that are using AI in a practical way. Informed.IQ is one of those companies. Informed.IQ unlocks the information trapped in loan applications and eliminates manual reviews via AI and machine learning.

The company works with financial institutions of all sizes to help them make the lending process more efficient and reliable. I really enjoyed the conversation as we delve into the benefits of AI, the differences between AI and robotic process automation or RPA and some career advice for aspiring fintech entrepreneurs. I hope you enjoy the show.

In this podcast you will learn:

  • How AI is helping lenders of all sizes
  • How AI is different from RPA
  • AI is still in its infancy
  • Complex scenarios still involve human intervention
  • AI should not be generalized
  • A new credit score is needed
  • True income is not being realized today
  • 15 – 20 percent of applications had falsified documents accompanying them during 2020
  • Covid was an accelerator for Informed
  • Read “Power Up: How Smart Women Win in the New Economy” by Magdalena Yesil, Informed’s Co-Founder
  • Justin’s parents did a good job raising him
  • And more…

Download a PDF of the transcription of Podcast No. 9 Justin Wickett

PITCHIT FINTECH STARTUPS PODCAST NO. 9–JUSTIN WICKETT

 

Welcome to PitchIt, the fintech startups podcast, one founder, one startup, one investor at a time. I’m your host, Todd Anderson, Chief Product Officer, LendIt Fintech.

(music)

Todd Anderson: On Episode 9, we talk with Justin Wickett of the Informed.IQ. Informed’s mission is to unlock the information trapped in loan applications and eliminate manual reviews of applications via AI and machine learning to enable touchless funding. You know, AI’s kind of been one of those most used buzzwords in fintech and, you know, I think Informed’s practical application of the technology is really helping to prove it out for most of the lenders that they work with.

Justine and I discussed how Informed’s not only helping lenders, but it’s also helping to catch fraudsters, the differences between AI and Robotic Process Automation or RPA and he also offers a career advice for those interested in getting into the fintech space. We had a lot of fun on the episode and I hope you all enjoy the show.

(music)

Todd: Welcome to the podcast, Justin, how are you?

Justin Wickett: I’m great, Todd, thank you so much for having me on this podcast.

Todd: Yeah, of course. So, first I’d like to, you know, just kind of kick off with telling the listeners a little bit about yourself, your professional background, kind of what brought you to where you are today at Informed.

 

Justin: Yeah. So, I’m Justin Wickett, I am CEO and Co-Founder of Informed.IQ. Informed is a software company that automates the consumer loan verification process for lenders using AI. Before I started this company, I was doing product management over at Credit Karma, did that for several years working on the auto refinance product over at Credit Karma as well as also their auto insurance product line.

Before that, I was doing product management at Lyft, the ride-sharing company in the very early days, really focused on consumer acquisition to passenger acquisition, passenger engagement and before that was at Zynga, the social gaming company. I have a background in Computer Science myself.

Todd: And so, what kind of brought you to the moment of starting Informed, what’s the founder story there. It sounds like you’ve worked for a variety of companies, have you always had kind of the bug to eventually start something for yourself, you and the Co-Founders.

Justin: Actually, I have. In college, dabbled in several entrepreneurial ideas, I’ve been busy in the early days of Twitter, had a Co-Founder and I over to talk about Twitter polling and search on Twitter, that was back when Twitter didn’t have a search engine, if you can imagine that. So, it’s always been in my blood, I’m blessed with incredible parents who have entrepreneurial background themselves. I started Informed based on some observations and learnings that I had during my time at Lyft and Credit Karma.

At Lyft, I got to watch so many drivers trying to sign up to drive on the Lyft platform and they didn’t have cars that could qualify. I watched these folks go out and I tried to obtain financing to purchase cars so that they could drive on the Lyft platform. Through that, that experience, I just saw how many documents would have to be exchanged as folks tried to prove out their income, prove out their residence, insurance situation, I watched the inherent bias in that process, just the lack of transparency.

When I shifted on over to Credit Karma that became even more just obvious, not just in the auto lending space but also for unsecured personal loans or other financial products seeing how financial institutions struggled with lack of software, these manual repetitious tasks to verify applicant information, I knew that there had to be a better way.

Todd: And so, if you could just give the listeners kind of an overview of kind of exactly the products that Informed offers and kind of what your ultimate customer that you guys target and work with.

Justin: Yeah. So, we sell to national financial institutions, those who are using our Software as a Service today, banks including Ally Bank, Westlake Financial, they’re leveraging this AI to fund the loans in real-time with greater accuracy when applicants cannot be verified through traditional data sources like Equifax’s the work number.

Today, when you go to get financing you oftentimes fill out a credit application and you make certain representations, you put down your name, your Social Security number so you represent who you are, you represent how much income you make, you represent where you live and for a small ticket loan maybe there’s no verifications required or minimal verifications required. But, as you begin to finance a vehicle or a home, property or assets there are a lot more verifications required associated with that loan and that’s in part for regulatory reasons.

The regulators want to make sure that financial institutions are not overextending credit, extending credit in manners that are predatory to folks that have no means of repaying those loans. Banks also have a real interest in ensuring that they’re able to get paid back and their fiduciary responsibility associated with managing their assets under management.

Todd: And so, you know, obviously anyone who’s in fintech today has heard of AI or, you know, it’s become quite the buzzword. I don’t know if it’s leading or at least up there with blockchain as buzzwords that everyone uses. As someone who runs a company using AI, you know, kind of where are we today and kind of the use of AI in finance and then how does that term and RPA, you know, the differences between the two, just kind of curious for the listeners to better understand as someone who uses this technology, where do you think we are today?

Justin: Yeah. I think the word is in its infancy. AI is really just getting started and there are so many applications for AI within a financial institution. We’ve chosen to focus on a very specific use case around loan origination, I mean, AI being deployed across financial institutions in the form of chat bots, in the form of fraud detection, underwriting, there is so much use of AI opportunities to automate. There is also a lot of functions within financial institutions that fit in the automation or are not quite yet ready to be automated, but our AI, in particular, is used to analyze credit applications and verify all supporting documents in an application file or loan jacket.

So, think of the promissory notes or retail installment sales contracts or title documents and appraisal guides as well as consumer permission data like sources like Plaid and Finicity, all that needs to be analyzed. The way I think about AI versus Robotic Process Automation or RPA is RPA is much more focused on automating a workflow, a process and it is stateless. By that, what I mean is it’s just repeating that process over and over again without any additional intelligence.

Todd: Got you.

Justin: AI introduces the intelligence that enables the predictions to actually be make. So, I think Gen One was really RPA being deployed across financial institutions, now what we’re really seeing is AI being infused into that RPA process to make intelligent predictions around say which documents are fraudulent, whether an applicant is really earning the income that they claim to earn, those are applications on AI.

Todd: So, in terms of where AI might be going and how much of our future could be automated, is there a fear that all of a sudden there’s this Robotic Age in a sense, not necessarily robots, but a lot of the things that potentially humans or some of the human intervention does today that there could be this massive wave of people then being let go…..I know I’ve heard the arguments of AI is it allows people to focus on higher end tasks, but over time, AI gets better, it gets smarter and do we then reach some sort of inflection point?

Justin: I think we’re a very long way away from that. I do think that AI today is enabling for folks to focus on higher order functions within financial institutions. For example, Informed automate each month the verification of over 100,000 applicants turning documents in data into decisions in seconds based on the financial institutions’ policies and procedures.

That work is very repetitious, laborious, manual task that a human just struggles with and it’s not fun to have to look through a hundred different bank statements each day trying to tally up all of the line item deposits so being able to free staff up at the financial institution to do a better job of providing customer service, doing a better job of interfacing with customers and bringing more loans in-house, driving conversion for the financial institution is incredible.

What Informed does is we automate the repetitious tasks, some tasks require a manual exception handling. The AI has not yet been trained to solve every single problem. There are so many sub-forms of income that an applicant might be claiming. So, for about 10% of the credit applications coming in and application files, Informed will require a manual review and there’s that manual exception handling process to handle particularly challenging situations such as say an applicant that is claiming income from multiple different sources on their bank statements.

Maybe they’re earning royalties and they also have others forms of income like some kind of child support court order where that child is turning 18 very soon so complex scenarios that require true human in the loop intervention is I think where we’re going to best serve folks that need help. When I call my bank, I oftentimes find myself sitting on hold for a very long time and I just wished the financial institutions did have more trained folks focused on serving the customers, focusing on these higher order initiatives. So, I think that’s very much what AI is enabling for today in financial institutions.

 

Todd: How much better is …a way from it’s use and application in fintech or financial services today, how much better is AI gotten the last few years, In guess, in terms of where the tech itself is going. Is it kind of getting better year by year, is it making jumps, leaps and bounds every year? Can you give the listeners a sense of how much smarter is it getting on kind of a year-over-year basis?

Justin: I think that the key to success with AI is focusing on very narrowly defined problem areas and doing an incredibly good job with that. I think that AI can’t be generalized, AI from company A is very different than AI from company B. The AI solutions that we see banks like Ally or CapitalOne and others in the market adopting are narrowly targeted, focused on a particular use case and can perform that use case with exceptional accuracy and speed and they require that the AI is proven through this model risk management processes to ascertain bias and precision and the stability of the models and measure the models’ drift.

So, AI I think is becoming much more targeted, it’s much more challenging to enter the market with an AI solution without having this massive trained data set that you can measure performance on. So, that’s the trend in the industry that we feel relates to AI and Informed’s been really riding that trend in the context of the consumer loan verification process for lenders.

 

Todd: You touched on it a few minutes ago, how much can AI and some of the stuff that you guys are doing increase access to financial services for people who might be shut out say by traditional FICO or something like that. You mentioned the Equifax work number, how much can it democratize beyond just say the prime, super prime, near prime type borrower?

Justin: Well, I think that’s really our opportunity, that’s what drew me to this problem and what motivates my team every single day. When I was at Lyft, I witnessed folks who were trying to apply to be drivers and keep in mind, Lyft employs more drivers than Fedex does on the road. I watched these individuals try to get access to credit and on their bank statements…when they submitted their bank statements to lenders, to car dealers, those banks statements had a bunch of ATM cash deposits on them and lenders, in some cases, didn’t treat that as income.

In other cases, they did treat it as income, sometimes the bank statements reflected deposits from Cash App or Venmo and oftentimes lenders would get that data, be it Plaid or Finicity, but they wouldn’t know whether or not that’s actually true recurring income and we realized that there really needed to be an AI solution to more accurately determine what is real recurring income so that we could democratize access to credit and remove bias from the loan origination process. So, there was a huge opportunity staring at us and as we begun to talk to more and more lenders, we realized that they really struggled with implementing their written policies and procedures, what they were representing to the rating agencies, to their regulators. If you talk to five different loan processors, they will give you five different answers as to what deposits actually are to be considered as real income.

So, we saw a huge opportunity to automate that process and ensure that ….say the hairdresser who is applying for a loan to buy a car and get to work, while they’ve got earnings on their pay stub, some of those earnings weren’t being considered…maybe some double time pay or some overtime pay, wasn’t being considered in certain cases or different regular pay line items weren’t being considered and in other cases if they were giving haircuts outside of their hair salon, the income that they were earning via Venmo, via Cash App, that too wasn’t being considered. We just saw that there was a better opportunity to enable for financial institutions to better recognize what is true income, leveraging this contributory data set across the industry.

Todd: It also plays to the shift in employment, generally, that people are now choosing a lifestyle and then finding work to fit a certain lifestyle versus I’m going to be in a 9 to 5, now I have to see the life around 9 to 5. That kind of equation has shifted and I think even more so in the pandemic where remote work has now become a thing that everyone’s essentially been forced to do that’s kind of thrown a lot of models on their heads.

Justin: Yeah. We really see that in the data when we analyze the 25+ million records that we have upon file this contributory database that we maintain for the industry. It really stands out to us, especially accelerated by the pandemic where consumers are having to work multiple jobs, folks are driving for Uber and Lyft and PostMates or renting their bedroom on Airbnb. People are earning income through more channels than ever before and financial institutions are really struggling to quantify which earnings they should be considering as true income and to comply with the Equal Credit Opportunity Act, to comply with Fair Lending laws. That is paramount for a regulated financial institution entity.

Todd: Do we need a…kind of an abstract question, but do we need kind of a new version of a FICO or credit score?

Justin: I think so. I think that there is apparent need for rethinking access to credit to ensure that more Americans are able to tap into the financial system. I think that FICO is limiting in many regards, same with also the VantageScore, there’s a huge opportunity to rethink access to credit in this country and we’re starting to see that trend ourselves with the lenders that we serve.

Todd: Do you guys…..just back to some of the stuff that you guys do, do you plans to kind of move beyond lending? What about small business lending, you know, have you I guess thought of that. To me, on the surface it seems like some of the stuff can be applied to small business lending. Obviously, data sources and stuff would need to apply to what the businesses do, but just kind of curious if you guys have thought about that.

Justin: Yeah, we have. We’re not quite there yet, but we’re on a journey to get there that’s why we recently raised $20 Million from some great firms, Nyca of New York, US Venture Partners, venture capital firms and for us, a lot of the small business lending is actually to individuals, individual principle and there is a lot of need to analyze documents like tax returns with Schedule C’s and bank statements. So, we do do some of that today, but there’s obviously a huge opportunity there. The models that we built, that we spend years training up to declassify, to extract, to compare information and this contributory data set that we’re maintaining for the industry also has applicability in other areas like health and human services.

We’ve actually been contacted by numerous government state agencies, Administering the Medicaid Program or supplemental nutrition assistance programs. Over 25% of the US population takes advantage of government-subsidized health care or these benefit programs and just like a bank who has to process a credit application, these data agencies have to process applications for these benefits and review all of the pay stubs and Social Security income award letters and military leave and earnings statements and the VA award letters. There’s so much documentation and data that needs to be analyzed that we can leverage this data processing pipeline we’ve created to bring more efficiencies and automate a lot that workflow.

Todd: Outside of some of the stuff we were talking about in terms of the credit scoring and stuff like that the last year, did you guys see any, I guess, you know, the fraud element. You know, there’s a lot of government capital coming into people, kind of how did you guys see a lot of that stuff, just kind of curious to hear how that worked the last year, especially with all the different data sources that you guys have. I mean, I’ve heard from others that there was basically billions taken out of either people’s pockets and stuff like that because there had to be so much money thrown out to people so quickly that it was kind of…..it got to a point where, you  know, fraud was an accepted piece of it though in reality it didn’t necessarily have to be that way.

Justin: Yeah, yeah. During the pandemic, we have had more and more financial institutions that never before reaching out to inform wanting to leverage the AI models that we’ve been training up to identify a lot of the fraudulent pay stubs or fake bank statements. If you do a Google search right now actually for fake pay stubs generator, there’s hundreds of thousands of (cross talking) that come back that let you create fake pay stubs, fake bank statements and it’s impossible for the staff of banks to be trained up on every single template out there on the dark web so……

Todd: You know, some of the theory behind that is they’re just going after some of the small ticket loans and banking on verification not really being that strong.

Justin: All of the above, we certainly see that. We are used by some of the biggest auto lenders in the country to identify fake documentation and abnormal income that applicants represent and we see it all the time. Normally, pre-pandemic, it was about 1% of the loan volume flowing through had these falsified documents in it. During the pandemic, for online lenders, that jumped up to 15/20+% of applications had these falsified documents accompanying them. So, it was incredible and I think that there is a very real need as financial institutions move online and neobanks process more and more online applications to be very vigilant and aware of Leverage AI to identify these known fraudulent patterns.

Todd: I saw recently that you guys were accepted into the FIS Fintech Accelerator Program, you know, can you tell us a little bit kind of how that went. What does getting into something like that mean for you, the firm, your employees?

Justin: Yeah. I think FIS….saw that we were literally saving up banks millions of dollars each month by helping them identify things like fake pay stubs, fake bank statements and FIS and the Venture Center saw real opportunity to select and partner with Informed and help Informed present its solutions to the community banks, the regional banks spread all across the United States to empower them with the same AI and machine learning tool sets that banks like Ally and Westlake Financial and others have access to.

So, the FIS Fintech Accelerator is a very prestigious program that we’re delighted to be part of. There were over a thousand different companies that applied, only ten were selected, we’re thrilled to be one of those ten and to be able to work closely with FIS and leverage their capabilities to serve the community banks and regional banks across the United States has been a privilege.

Todd: I want shift a little bit kind of how you, the team have been, you coping the last year and then we’ll get into a couple of other things about kind of career advice and the fund raise, but kind of the last year, how has it been, you, on the team? Are you guys fully remote, do you have plans to kind of do hybrid, go back to the office? You know, I’ve heard interviewing founders, there’s a variety of different models kind of creeping up now that people have started to get a little bit back to normal, how’s everything been for you and the team?

Justin: Yeah. It’s been incredibly productive for our team, we’re blessed in that software engineers, we can work very effectively in a remote environment. That’s not so much the case for our customers, our banks realized that their staff who previously were having to manually look at documents on desktop computers in the corporate office no longer could access these sensitive information and banks had to issue laptops to their staff.

You can imagine some bank employee looking at your tax returns, your pay stubs, your bank statements in their living room with a let’s say roommate walking by sneaking a peek. Banks realized that they really needed to leverage AI to automate this function to minimize the number of humans required in the loop here to review this kind of sensitive information. So, we have worked very well remotely, we have plans to go back into an office environment, we will continue to work remotely developing these machine learning models. For Informed, COVID has been a real accelerator, it really drove the adoption of more and more AI because of some of these reasons we just mentioned.

Todd: At any point in your journey, have you gotten close to saying I’m not sure this is going to work, you know, maybe this was a good idea, but might be too hard or have you been kind of driven. You know, we see this solution working and we need to stick to it.

Justin: No, quite the opposite. We kept hearing from our banks about this problem, this was being asked of us, hey, can you help us with this huge problem that we have analyzing all of these data that we’re getting from these applicants across the United States. Initially, I mean, it’s a daunting problem, how can you process millions upon millions of documents, but we’ve got a great team of software engineers, we really put our nose to the grindstone, we wanted to do the best job possible because what’s at stake here is some Americans access to credit and ultimately the calculation of income is so important for debt to income, payment to income calculations, part of an underwriting process.

It literally is a determining factor for what kind of interest rate you’re going to get so we  saw this opportunity, our banks kept begging us to invest in it to help them with it and really aggregate data across the entire lending ecosystem. They wanted to be able to ensure that they were harnessing best practices, say a new fraudulent pay stub template emerge that another bank was getting hit hard with, they wanted to make sure that if all of a sudden they started getting a flood of applications with that kind of falsified documentation that they would be able to take advantage of it.

So, this has been, at every turn along the way, clear to us that we need to be doubling down on this opportunity, otherwise, someone else will do it. If we don’t serve these financial institutions, there will be another faster startup that emerges that does a better job of bringing more efficiencies to these financial institutions.

Todd: You mentioned before the recent fundraise, you know, how did that go for you guys, obviously, you’ve got some great names to come on board. You know, for the startup listening and, you know, this appearance that raising money today is easy, you hear stories in TechCrunch, Business Insider, all these stuff, tell us a little bit about how that went. Obviously, it’s not as easy as it appears even as….you know, we’re in the frothiest of times, how did that go for you and the team?

Justin. Yeah. For every company that gets funding there’s probably 50 that don’t. It is always a challenge, it always is a distraction taking you away from serving your customers. I think what investors saw in Informed is the massive financial institutions that we serve today has a real success metric and our ability to deliver real return on investment to these financial institutions that really transformed a process for the banks enabled for greater transparency to a consumer’s financial options in real-time so that is what drew great firms to invest in Informed.

Our process was relatively quick, actually we’re oversubscribed, we ended up having to turn some capital away, but that’s not always the case. Fundraising is incredibly challenging and daunting and any founder that goes through it and is successful is a true testimony to what their team has created. I think that…..yeah, so many companies fail in that process.

 

Todd: For the listeners, do you have any career advice and kind of those that are looking to get into either project, product management, fintech, it’s a real exciting time to get into the space, there’s probably more options now than ever before with remote work being, you know, still something that’s going to happen, at least in the foreseeable future and so talent can be acquired almost anywhere now. So, what’s some of your advice for any of those listening that are looking to get into the space?

Justin: Yeah. I think that coming by really great talent is so challenging for companies. Informed is hiring across the board, we are going to more than double the size of our staff this year and, again, in 2022 so we are actively hiring all across the board. I think that for someone who wants to get into product management or break into fintech, joining a company that already has product market fit, that already has customers that can actually provide you with very clear requirements is a true gift and shouldn’t at all be underestimated. We’re so fortunate at Informed to have that.

What drew me to Credit Karma was, again, the ability to work very closely with financial institutions and really listen and observe the workflows that they had in place and gather requirements, gather an understanding of how software can be used to better serve consumers applying for credit. Getting a loan sucks, it’s a horrific process and if you can study that process because they don’t teach it in school, if you can actually study the process by observing what lenders, big and small, are doing today, I think that you have the opportunity to someday start a company on your own because you’ve got that deep solid understanding that few people actually have.

Todd: You know, we’re getting towards the end of our time here. I usually end it with some fun. Curious, do you have a favorite book and the last book you read. If you don’t consume information by reading then choose the medium of your choice.

Justin: Yeah. I do actually have a favorite book and it actually happens to be the last book that I read. It’s my Co-Founder’s book, Magdalena Yesil published a book recently called “Power Up: How Smart Women Win in the New Economy.” Magdalena is a serial entrepreneur……

Todd: She’s great.

Justin: …..her fourth company so that book is a joy for me to read and oftentimes, find myself, looking back. Not only does it have great advice for women really kick starting their career, but also for men. I’ve truly enjoyed that book and it is one that I always go back to.

Todd: Are you a sports fan and if you are, do you have a favorite sport and teams that you root for?

Justin: Oh, man, I think that the easy answer there is Duke basketball. I went to Duke University…..

Todd: There you go.

Justin:…..and I got a chance to watch Coach K coach, he’s an incredible leader and Duke has always done incredibly well with basketball and I’m thrilled that they’ve got Jon Scheyer coming in as their new coach.

Todd: Yeah. I’ve heard big news of Coach K announcing that, big news.

Justin: Yeah, I kind of right after Roy Williams.

Todd: Last question, biggest inspiration in life.

Justin: I’m so blessed to have folks that have mentored me, who have invested time in me. I would say that both my parents are my biggest inspiration, but I wouldn’t be where I’m at today if it wasn’t for elementary school teachers and middle school teachers that took time to coach me, to give me feedback and invest in me along the way. So, yeah, truly so many folks that I would love to thank, my parents, in particular, I think, Mom, Dad, you did a good job raising me.

Todd: With that we’ll end it. Justin, I want to thank you for coming on the podcast, thought it was a of fun. Wish you and the Informed team continued success and we’ll get you back sometime in the future.

Justin: Hey, Todd, thank you so much for taking the time.

Todd:  Of course.

(music)

Thank you for listening to the latest episode of PitchIt, the fintech startups podcast. I encourage you to take a few minutes to write a review or rate the episode. Ratings and reviews both help us to improve the show for future episodes.

If you’re interested in learning more or would like to be considered for a future episode, please reach out anytime to todd@lendit.com and until next time.

(music)

[/expand]

You can subscribe to the PitchIt: the fintech startups podcast via Apple or Spotify. To listen to this podcast episode there is an audio player directly below or you can download the MP3 file here.

Subscribe
Notify of
0 Comments
Inline Feedbacks
View all comments