Tom Wilkins of Mortgage Media spoke with Steve Butler, Founder and General Manager of AI Foundry, at the California MBA Innovators Conference in San Diego, and discussed how AI and Machine Learning will disrupt the mortgage industry.
Butler, who is leading AI Foundry to deliver software solutions that incorporate artificial intelligence technology, has vast executive experience and leadership at both public and private companies in Boston and the Silicon Valley. He’s spent much of his career involved in different industries (chip design, semiconductor design, and enterprise software), but there has always been a common theme – process automation.
“In terms of financial services, I saw mortgage as an area of just very inefficient, and one in which we could get halfway there and still save a ton of money for companies. And if we could get all the way there, we could disrupt the industry completely,” Butler said. “Overall, my strategy is to bring a very disruptive process automation to the mortgage industry.”
Solving Real Problems
He noted that it’s important to attend regional conferences because he meets people looking for real answers to the technology problems they are facing. The people he encounters are desperate to solve very specific problems that will save them a lot of time and money.
“We have that depth of conversation,” Butler said. “I’m in requirements gathering mode when I’m here. I’m selling, but I’m really more about gathering requirements, and I take that back and it turns into roadmap for us.”
He pointed out that much of the interest from the mortgage space is around Artificial Intelligence (AI) and Machine Learning (ML) and how those technologies can change the mortgage business.
“I’ve been around long enough to know that AI had a renaissance. Well, it’s in a renaissance now, but it had an initial start back in the 70s and 80s, and it kind of flamed out.” Butler said. “It over promised and under delivered. But we’re in a different era now, and I call it a renaissance because, finally, real use cases have shown up that have delivered. Facial recognition. We trust our personal information on our phones because we have a facial recognition app that makes sure it’s us. We trust autonomous cars.”
The idea, he said, it to leverage AI and ML so it can be applied to the mortgage space to the point where it’s smart enough to replace humans doing the manual work of sorting documents, and classifying and extracting data.
“That’s what we’re going to deliver and I think it’s a breakthrough, and it’s going to deal with a big problem, which is that the cost to process a mortgage is going through the roof,” he said. “If you can put machines at work to do it, you’re going to solve it the same way cars are now robotically manufactured and humans are not involved. You’re going to solve it the same way here.”
AI is Here Now
The premiere selling point for AI for mortgage professionals is eliminating their greatest pain point – dealing with thousands and thousands of documents.
“You start looking at all the state regulations and all the other documents, bank statements that can be from a thousand banks. That’s a problem. It’s a huge problem. It’s a manual problem and a lot of labor to sort and do that. We can solve that today,” he said.
But the challenge to overcome that adoption hurdle, according to Butler, is that many mortgage companies and banks don’t actually understand their process at the detail level.
“You can’t automate anything if you don’t understand the process,” he said. “It’s still too much of an art form. It’s still too much in the minds of a few people. You go to a lot of these smaller businesses and they’re still doing several thousand mortgages a year, and they’ll say, “Well, you need to talk to Sally. She knows how the whole thing runs.” And if Sally’s not there that day, no one knows how it runs.”
To change that process, businesses need to document the process at a very low level flow diagram, where the data flows, all the documents flow, how the workflow works. And only then can you move forward, he said.
“Everybody’s conservative about implementing, so that’s an issue. You have to overcome a conservative nature. But I think people are getting that they have to change, but to change, they have to understand what they have.”
Change is on the Horizon
Once that mind shift happens, which Butler predicts will be in a couple of years, AI will be pervasive is all areas of mortgage – both consumer facing and the back office.
“Frankly, I see a time, and it’s not that long away, maybe a couple of years, in which the back office will have zero humans. There’ll be over oversight, but you’ll send a mortgage and out, it’ll be a completely automated process.,” he noted.
And that will be dramatic and disruptive and will lead to a shift in jobs for people were the process is fully automated, but other gains will be more beneficial.
“I also think the volumes are going to go way up because if you can turn the cost, and bring the cost back to the consumer, people will think about getting in and out of mortgages maybe every year.” he said. “Today, we look at it as a few times in our lives, but if the product can turn quicker, if you can get a mortgage in a week or a day, and the cost is reasonable, you’ll trade up off of a quarter point or a half a point change in mortgage rates. So I think it’s going to change everything if you can turn it around quicker.”