Recently Informed CEO, Justin Wickett, sat down for a chat with Peter Renton on the Fintech One on One Podcast to talk about how Informed’s AI technology helps lenders make accurate lending decisions by automating verifications in real-time with 99% accuracy. They also explored:
- How the automated verification process works.
- How the gig economy has complicated income verification.
- How informed uses AI and Robotic Process Automation.
- What the company is doing to prevent fraud
- Why Informed started with auto loans. And,
- What the partnership with Origence (a CU Direct brand) means.
You can listen to the whole episode here.
Some snippets from their conversation:
Peter: Tell us a little bit about why you started the company.
Justin: I was in Product Management at Lyft and saw drivers struggle to join the platform because they didn’t have a qualified vehicle. So they had to get a loan to purchase a car and I saw how broken the underwriting process was for a lot of these people. Many didn’t qualify for the 0% APR you sometimes see or lease deals you see advertised. These people would spend a whole day at the dealership trying to qualify for a loan and they weren’t sure if the lender would count their overtime pay or tips. It was a broken process.
Peter: Okay, I’d love to get a solid description of how it works.
Justin: Informed automates verifications in real time for over 150,000 Americans a month using machine learning and artificial intelligence. That automation reduces both manual keying errors and bias – which is really important, especially to full spectrum borrowers. So we automate the back office of a bank, and typically integrate with the loan origination system (LOS). We’re the AI that replaces the verification screen within the LOS.
Peter: Okay, so let’s dig into that. Can you provide an example of some of the documents and the data you’re looking at? Kind of take us through the process?
Justin: Absolutely, this is what fires up our team because though we’re a b2b company, we’re very mission driven. We talk about examples of, how a car dealer was trying to get paid by a bank for some kind of warranty or insurance product, that the car buyer never agreed to purchase. Our software goes through the complex contracts lenders receive from a dealer. This includes documents like Retail Installment Sales Contracts (RISCs), ancillary product contracts, vehicle service contracts, prepaid maintenance plans, etc. There are over 8,500 different variations of these contracts in addition to additional documents. We process documents so lenders can lend with greater confidence, avoid consent decrees, and better adhere to the policies required by regulators and rating agencies.
We process 10s of millions of consumer documents like pay stubs, SSI award letters, bank statements, W2s,1099s etc. And we see that about 30% of people applying for credit actually understate their income, especially in the nonprime consumer base. And the question is why would you understate your income? You’re going to end up with a higher interest rate! And the reality is, we’ve conducted user interviews and folks say, I’m an hourly worker and I don’t know how a lender is going to count my overtime pay or commissions or tips. I don’t know if I should include them. So it’s confusing and as a result, we see instances where a dealer will actually ratchet up someone’s income and the borrower may end up in a car they can’t actually afford. We also see dealers that struggle to figure out how the lender is going to calculate that applicant’s income and they understate it. As a result, they might not be able to help that applicant finance a vehicle that they really need. So that’s the situation we set out to improve.
You can hear more of the podcast here.
We are digital transformers of tedious manual tasks, bringing robotics process automation to financial services.