21 hours ago
8
Mortgage lending has long catered to a bygone era of steady paychecks and W-2 forms, leaving millions of self-employed and gig economy workers underserved. But as the gig economy booms, lenders are finding success in the nonqualified mortgage (non-QM) sector by engaging with AI-powered tools like Prudent AI.
“Our technology eliminates the tedious ‘stare and compare’ work of analyzing bank statements while enhancing fraud detection and ensuring consistency across underwriting decisions,” said Paul Gigliotti, chief growth officer at Prudent AI.
“The results speak for themselves: Processing times reduced from hours to minutes, productivity increases of 4.5x, and underwriters who can focus on what they do best —applying their expertise where it truly matters.”
HousingWire spoke with executives at three companies — Sierra Pacific Mortgage, Logan Finance Corp. and Angel Oak — who all use Prudent AI for streamlining their operations and encouraging growth in the non-QM space.
These interviews have been edited for length and clarity.
Sarah Wolak: How has the implementation of Prudent AI impacted productivity in areas such as underwriting and processing?
Suzy Lindblom, chief operating officer, Sierra Pacific Mortgage: We’ve seen a tremendous productivity lift. We’re still early in the process. Underwriters are still getting used to that product, but yes, I have seen an improvement in productivity. One example is that it allows us to do income calculation upfront for our broker partners and our loan officers, which helps them better prequalify the borrowers.
Travis LaLonde, executive vice president of credit, Angel Oak: Underwriting productivity improved by 4x. Full manual review underwriters may do one new underwrite on a daily basis because of the review, especially if you’re doing a 24-month bank-statement review.
That’s a considerable amount of paper to go through. With Prudent AI, we are absolutely more productive and we can touch more files on a day-to-day basis. The quality has been far higher. We’re doing it with a lot more efficiency, but it is definitely a partnership process.
Wolak: How does artificial intelligence help navigate the often-complex financial structures?
Paul Jones, senior vice president of business development, Logan Finance: In finance, we evaluate individuals’ bank statements for calculating income. With non-QM particularly, the whole idea is to take the work out of the questionable loan as quickly as possible so that the rest of the loan feels exactly like that originator or lender is used to.
With Prudent AI being able to give us a solution that delivers clear and reliable results, it really helps. One of the other things that we leverage with those results is using them to educate our customers better to understand the bank-statement loans or other income calculations. … Speed and accuracy of execution is what we use AI for the most. The quicker someone receives a reliable response from this platform who needs this type of program that we offer, the more likely they are to win the deal themselves.
LaLonde: I think a big piece for the AI for me — and especially our work with Prudent AI — is the fraud detection aspect. … Not only are we trying to qualify the borrower, but we need to be very cognizant of any red flags or anything that we’ve seen that may point to alterations having been made — which, unfortunately, in this day and age, it is a lot easier to manipulate certain documents.
Without having to get an old-school calculator out and add totals and make sure deposits and withdrawals are all calculated appropriately, that’s where that’s been huge with Prudent AI. It will call out those red flags, month-to-month statements that just aren’t computing properly, where it appears line items may have been altered, things along those lines.
Lindblom: Since implementing Prudent AI, I don’t have individual underwriters doing their own interpretation of income. I have consistency in looking at bank statements and looking at the income in a certain way.
The other operational advantage is the time that it takes. For an underwriter to look at 12 or 24 months of bank statements, you have to stare and compare. You have to look at individual bank statements and hope that you don’t miss anything. With Prudent AI, I have gone from about an hour to review 12 months of statements to 15 to 20 minutes.
Wolak: What impact has AI had on your ability to compete with lenders in the hard money sector?
Jones: The way to help loan officers is to make it very easy from our side. We don’t require them to register a loan with Logan for us to run income calculations.
The way we position it is, if someone is shopping around for a home and they’re self-employed, and they already know that their tax returns won’t work for qualifying, but they’re still not sure how much income they have, they can start with us. We can give them a number that they can put into a prequalification for the real estate agent and shop around now.
If it’s $10,000 a month in income, they know they can get full underwriting and get them approved, but at least we can do that upfront and those loan officers can get ahead of that game.
Lindblom: AI is our assistant. It’s giving everyone an assistant, helping to do the tasks that I don’t like to do, like research and the “stare and compare” work.
Balancing AI with human intervention is critical, especially to verify the facts and to make it personal. A lot of organizations may use AI but then they don’t verify — there is still a need for human verification. For instance, our underwriters still look at the Prudent AI output to make sure that everything is correct and consistent.
English (US) ·