One-On-One: CEO of Dark Matter Technologies Talks About Transition, Automation

The CEO of Dark Matter Technologies says they remain focused on streamlining processes for their clients as they work to revolutionize the mortgage industry.

Leaders at the company, formerly Black Knight Origination Technologies, announced in September that they had rebranded after being acquired by the Perseus Operating Group of Constellation Software.

Since that time, they have announced a collaboration with Argyle to simplify the income and employment verification process and an integration that allows lenders to instantly leverage income and employment information through Experian Verify.

They have also implemented an automated workflow in the Empower loan origination system that makes it easier for lenders to identify and qualify loan applicants who could benefit from an evaluation of their rent payment history.

Rich Gagliano recently sat down with The Mortgage Note Editor Kimberley Haas to talk more in-depth about the transition and the technology that is available to today’s mortgage professionals.

Haas: So Rich, we haven’t had an opportunity to sit down and chat yet. I’m hoping that maybe you could start by telling us just a little bit about yourself.

Gagliano: I am originally from St. Louis, Missouri, and I live down in Jacksonville, which is where we’re headquartered here at Dark Matter. I started out as an underwriter in the mortgage space.

I came here about 13 years ago to help with the product and that evolved into being the president of the business as part of Black Knight. And then as the acquisition was happening and we got carved out, I assumed the CEO role.

Haas: How has that transition been going?

Gagliano: We were part of a bigger organization, so we have some work to do around setting up some systems, that sort of thing, back office. But a lot of that work’s already been done.

We’re switching out some platforms that we think are better for us versus enterprise-type capabilities. And we’re keeping close to our clients.

So the transition’s been going really well. A lot of just administrative stuff and letting the team focus on taking care of our customers, getting them implemented, those sorts of things.

Haas: Now, what kind of clients are you working with? Has that changed at all?

Gagliano: I think the market, as you know, has been really tough. And we’re seeing less independent mortgage bankers right now. We’ve seen more contracts from banks and credit unions.

Haas: I know that a lot of your mission is to revolutionize the mortgage space using technology. Can you tell us how artificial intelligence and machine learning are helping with those efforts?

Gagliano: We’ve been on this mission for a while, and we have what we call our orchestration engine, but it essentially, it’s automated workflow, and it’s robust and it does a tremendous amount.

About six, seven, years ago, we realized that we were kind of hitting a point where we couldn’t advance it anymore without some other capabilities. And so we started down the AI path.

What that’s doing for us is it’s allowing us to reduce and eliminate a lot of the manual processing going on.

When borrowers are submitting their documents, you don’t need a human to go intercept it and push it into it’s electronic folder. We do that with technology. We pick the document up, we look at it, we understand what it is, AIVA does, which is our artificial intelligence.

So she’ll look at it, understand what it is, and puts it in the right place. And if it’s a doc that we need data from, we pull the data off, and then we run rules against it. For example, we compare it to what you just told us your income is. And so we can do the math for the lender, and we’ve just eliminated a lot of steps.

We’re looking for not only the pulling the data off and doing the calculations, we’re making sure you’ve submitted the right information.

We asked for two W-2s, we checked and made sure that they were two right W-2s. And we needed 30 days worth of pay stubs. We checked that, and if you forgot to do it, we alert the user upfront. So we don’t wait two weeks, three weeks, for the underwriter to realize that they’ve got a wrong document.

We call that exception-based processing. We get back to the processor or user so they can get back with the customer.

And we’ve extended it a little bit further, and what’s coming out is we’re actually communicating with the borrower through technology.

We’ll ask you as the processor, “Would you like for me to send a communication to the borrower?” And then we essentially send a communication, “Come look at the portal. We’ve got some information for you and we’re missing a pay stub, or we need this W-2.”

If they upload it, we evaluate it, and we check the box that we’re good.

So there’s a lot of automation going into what we’re doing to help that lender drive down that cost, and our view is if we can drive down the cost, we give that lender a competitive advantage.

Haas: What are some of the other benefits to using AI? I’m sure that there’s enhanced accuracy.

Gagliano: AIVA is looking at documents and pulling data off of documents. There’s not too much you can submit that she doesn’t know about, and can handle, and can pull the information off and do the testing, do the checking. The accuracy is in the 99th percentile. So it’s actually better than a human. Humans actually will make more mistakes moving documents around than what the technology will do.

Haas: I’m sure there are many ethical concerns that come into this space. What are some of the key things that you are thinking about and that your colleagues are thinking about as this technology is rapidly improving?

Gagliano: We stay focused on where we can streamline but are not getting involved with doing automated AI underwriting. The industry’s not ready for it. I don’t think AI’s ready for it.

It has to be able to explain why it approved you and didn’t approve me. Even if it ended up being the right answer, it has to be able to explain that. And it’s just not quite there yet. So no focus there.

Our view is, and my view is, that we are there and we’re using the term assistance. We’re creating these assistants to help that user, processor, closer, and underwriter. But they’re making the determinations.

We’re bringing information and showing them the calculations and helping them get through these loans faster. But ultimately, it has to be the underwriter that makes that decision, in my view.

You know, back when I was an underwriter 35 years ago, we were doing 15 underwrites a day, and there was no technology. It was all paper. But the world’s changed.

The amount of scrutiny, the compliance that has come in over the years, it’s burdensome for these underwriters to get through this work.

We believe that we’re going to get underwriters, we’re trying to be conservative about it, but we think they’re going to get them to about, 10 a day.

The big question that we got as part of this whole process is, “Are you guys still going to invest? Are you still going be thinking the same way? Are you still going to transform the industry?” And the ownership has been great and absolutely in support of what we’re wanting to do.

They get it. They understand it, and they know that to our clients, it’s important to them. They came out and met with quite a few clients with us, so that they can hear it from them too. They’re in this for the long haul. They’re a buy-and-hold.

They want to make sure they take care of the customers and they are being supportive. And that’s our vision.

Editor’s Note: This article was originally published on March 12 in The Mortgage Note. See the original here.