Optimal Blue’s Erin Wester on using AI to meet the challenges of this market

Name: Optimal Blue’s Erin Wester on using AI to meet the challenges of this market
Uploaded: 2026-10-07T08:00:37+00:00
Duration: 19 min 36 s
Channel: HousingWire

00:10Welcome, everyone. My guest for this sponsored episode is Erin Wester, Chief Product Officer

00:15at Optimal Blue, to talk about how Optimal Blue's continued investment in AI, including

00:20their virtual economist and the launch of AI Labs, helps lenders meet the challenges

00:25of this current market. I'm excited to get started. So Erin, welcome back to the podcast.

00:30Thank you. I am very glad to be back. Thank you for having me.

00:33You know, I love talking to you, whether it's on stage or on the podcast. So we've always

00:38had some great conversations. This one's going to be really interesting, right? So, you know,

00:43the industry is facing a major turning point January 1st. So, you know, FHA will allow Vantage

00:50Score 4.0 and FICO Score 10T alongside Classic FICO for eligible loans, which is a big change.

00:57But Optimal Blue has already been ready for a while now. So I'd love to, you know, get

01:02your take on how your team has been preparing for industry adoption of these new credit models.

01:08Yeah. What a time to be alive in the mortgage space, right? It's a very big, very big change,

01:13very exciting change. I think this will be really telling into how kind of shaking up things that we

01:19have relied upon for decades, how that could help with affordability. So I'm very excited to see what

01:24the future has in store, especially that post-January date. We've been, you know, hard at

01:29work getting the credit score model changes into our platform, into our engine. But we've really

01:35tried to be diligent about ensuring that it's not just how Optimal Blue will be supporting these new

01:42fields to support the rate sheets and, you know, support any of the LPA grids that could come out

01:48or have come out and really make sure that we have, you know, the lender's best interest at heart

01:53when doing that. So making sure that how we build this out works as seamlessly as possible through a

01:59lender's tech stack. And that's been some of the challenging pieces of making sure that we've

02:03been communicating clearly. We understand how our loan origination system partners will support this.

02:09Our point of sale partners will support this. And making sure that that data flow from start to

02:14finish for our customers is as seamless as possible because it's critical. Credit score,

02:20credit modeling is critical from pricing, processing, underwriting, you know, all the way through.

02:25There's not much more you can rely on, you know, that critically. So making sure that it's flowing

02:32properly and then making sure that, of course, we're evaluating it correctly and giving lender choice,

02:39giving them the ability to run side-by-side scenarios as folks start to learn

02:44which model will fit which borrower scenarios better as people get more reps under their belt

02:49was another really big important facet of how we implemented this, making sure that

02:54we allowed for tools in our originator assistant that surfaced those kinds of suggestions,

03:01making sure that we're putting our, you know, customers first and the fact that we're trying to

03:07make sure that the best price product comes out for the borrower at the end of the day.

03:10So it's been a whirlwind, but it's been fun. And yes, we got that released earlier this year and

03:16glad to have that in our rear view. And now we just get to help customers adopt.

03:20No, that's amazing. You know, I would think that it's a whole different thing than

03:25for, you know, Fannie, Freddie, FHA, whoever to say like, okay, here are the new guidelines and you

03:30can only do this. And then now everything's that as opposed to giving people a choice.

03:34So from a software standpoint, from a lender standpoint, having that choice, being able

03:40to, you know, then it's actually, to me, more complicated. It's better, right? We all want

03:45more choices, but like, that's a more complicated thing than just making a change.

03:49Right. Exactly. Especially as, you know, people are navigating how they're going to implement this

03:54throughout, you know, throughout the organization. So how is this going to impact credit ordering in

03:58general? How is this going to impact cost of loan, both with the cost of the reports, but also the

04:04offsets and the LLPAs? There's lots of economics to this as well. So choice is very important because

04:11it's also going to act as basically like an education tool in the beginning while everybody

04:15gets their sea legs under them. When understanding, you know, is this a better profile for a Vantage

04:21loan? Is this a better profile for a FICO 10T? Is this a better profile for classic FICO?

04:26No, that's incredible. You know, I think that this just goes into the larger thing that Optimal

04:31Blue is doing, which is like, you're continuing to make investment in AI capabilities and then figure

04:36out the best use of that capabilities for your clients, for your customers, right? Like AI by

04:42itself, we all know now, like, it's great, it's fine, but like, it has to be leveraged, it has to

04:47be harnessed into the best use case. Otherwise, it could be, you know, not as helpful as it could

04:54be. So tell me a little bit about how Optimal Blue is investing in this AI capabilities.

05:00Yeah, yeah, well said, you know, AI's sake for, for building AI for AI's sake actually helps,

05:06I think, a lot internally, like on, on some things that go on within like the four walls of,

05:11you know, our tech organization. There is a lot that people are getting very, very creative with

05:16to help streamline internal processes. But when it comes to building something that customers will

05:22use, it really does have to solve a disparate problem, or better yet, solve a problem nobody

05:27even knew that they had. And that's, I think, where the real magic happens. And, and all of that

05:33requires, you know, dedicated focus, not only on the implementation, on how these things can work,

05:40how we measure the impact and utilization of those things. So from Optimal Blue's perspective,

05:46we've launched our AI labs that allows us to set a dedicated team who live and breathe this and the

05:52tech technology landscape of AI, which is a very broad term now, as we all know, changes daily. So

05:59it really does require that kind of dedicated focus and learning. And this run today by Kevin Foley on my

06:06team, and he does a fantastic job of keeping myself and the company all up to date on the latest

06:11and

06:12greatest. He's got a real passion for it, which helps drive good products, be that, you know,

06:17internally to make things more efficient, to get things done faster for our customers,

06:20or externally, things that we can present to our customers.

06:24So what does that AI labs look like? Is that something that people, you know, are,

06:29do you have groups in there talking to him? Is he like, how does it all work? And how do

06:34you get

06:34that innovation through that?

06:36Yes, he's got a lot of nicknames. So he's like our AI ambassador here at Optimal Blue.

06:41So he works with every department, every, every single department, and then can hear directly

06:46from our client facing departments on what they're hearing from our client base here from our investor

06:52content management teams on how he can help kind of streamline that critical part of our, of our

06:57workflow. But the team is structured with, with Kevin kind of at the helm and we have some machine

07:03learning developers. They partner closely with our development teams that support our core product

07:10lines that introduce things like generative AI into our platforms. So he's got a specific team in the

07:17lab, but really works with every single department as like our AI ambassador across Optimal Blue.

07:24And customers will, we'll be hearing more and more from him as well.

07:27That's so exciting. You know, we're implementing a lot of things in AI in our processes at Housing Wire.

07:33And I feel like every week there are new capabilities of what we can do. And so it makes so

07:39much sense to

07:40have somebody right there who's, who's keeping track of all that and also knows, okay, this is what

07:45people need. Okay. How do I take what I know can actually be done and marry those two together?

07:50Because if you didn't have someone doing that, I mean, people don't even know what's possible.

07:54So like having someone like him, having AI labs lets you go, this is what we're trying to solve

07:59for and let someone who can really marshal everything, figure out how to get it done.

08:05Exactly. And, and no idea is a bad idea. So we, we are encouraging, you know, a constant flow of

08:11communication and kind of helping write, guess this fits, or maybe this fits better over here. And you're

08:17exactly right. Not only do things change so rapidly on, on what it can support now, but also day over

08:23day, it just gets so much better at analyzing data. You mean, I don't know if you use a co

08:29-pilot

08:29for example, at housing wire, but just seeing what it's done to be able to produce PowerPoint

08:34presentations over the course of the last call at 30 days is, you know, monumentally better. So it's,

08:39you've got to keep a pulse on current things that, you know, it can do at getting better at it.

08:43And then

08:44also just net new technologies that come into the space, you know, with the, the agentic AI world

08:48changes every day. It is it's very much a world for an academic, which is great because Kevin's

08:55got a lot of that in his, in his history as well. You know, Optimal Blue was obviously part of

09:01our

09:01AI summit that, that we had in August. And one of the things that really stood out to me is

09:05that

09:07you're the people that you're trying to help the lenders, like they have a whole other job,

09:12right? And so they can't be the ones who are, who are keeping up with all of that, but they

09:16have to

09:17know and they have to trust that they're the people they're looking for to tech are absolutely

09:22at the cutting edge. Otherwise they're at a disadvantage. Completely agree. Completely agree.

09:27And I think it's the job, not only of tech providers to be staying at the forefront of

09:34understanding everything that's available at our fingertips, but it's also our responsibility

09:39to ensure that our customers feel educated and comfortable and confident with what we're producing

09:45and providing at the end of the day, because, uh, the, the adoption of AI allows for more stronger

09:54utilization and stronger continued development on top of that. So think about it like laying the

09:59foundation of a home, the more that you're able to kind of bring that into life and the more bricks

10:04you can layer on top of that, you know, the, the stronger and better your infrastructure becomes,

10:08be that processing alone or pricing alone or selling alone. Um, and I think, you know,

10:13we've been trying to be very, very, uh, discreet with how we implement these things to make them

10:19a part of the natural workflow that our customers are used to. So the adoption and utilization feels

10:24natural. And then that will build that confidence and trust, uh, because they're seeing it alongside

10:29of their day to day activities that they've always done. And really we're going to be putting kind of

10:33gasoline on that to, um, ignite more and more features that build off of that foundation to help kind of

10:40stretch everybody's, uh, understanding and everybody's comfortability to kind of go into

10:46that next space. And you'll see a lot of that actually announced at our summit, uh, here in

10:50February. Oh, I can't wait. Okay. We're going to talk about the summit, but one thing I wanted to

10:56touch on is, um, the way that you guys are utilizing AI is, um, you're a virtual economist,

11:01right? So you launched that, um, at, at your last conference, but like now you've launched general

11:06availability for it. And I would love to know, you know, how's it helping lenders to plan for

11:11what's next and maybe first tell what, what it is. Sure. Sure. Yes. We launched our virtual

11:18economist at our summit last year, as you mentioned, and just had general availability come in,

11:23in this summer. Um, we had a beta group that helped us in between those time periods really

11:28make sure, uh, that, that everything was functioning as, as the way that we had designed. And, uh,

11:34the virtual economist has been, you know, I think received very, very well from our customer base.

11:39And what our economist does is it, uh, aggregates proprietary lock volume data from our customers

11:47and marries that with forecast data available through industry publications, like the MBA,

11:52for example, and produces a proprietary machine learning algorithm that allows our customers to

11:58engage with an avatar gauge with the economist and ask questions about forecasting the rate

12:04market. Um, and then based off of those rate market forecasts, what they can expect to happen to

12:11their, their volume. We've got, you know, a lot of history for our customers and lock volume in our

12:17very cyclical and seasonal, even though maybe over the course of the last couple of years,

12:21doesn't feel that way, but a very cyclical and seasonal, uh, invite rate environment to be able to

12:26make sure that we have, you know, sound and sturdy predictions. Um, and I think, you know,

12:32maybe one of my, my favorite quotes from the summit last year, um, was, uh, was from the NBA,

12:38actually, when they came on saying, we agree with the optimal blue virtual economist. That was

12:41wonderful to hear. Um, and to, in terms of, you know, predicting the rate market throughout 2026.

12:46Um, so we've, uh, really it's about seeing around the corner, uh, making sure that our customers feel

12:52like they have a really grounded source to be able to plan for the future based off of different rate

13:00curves that we may be seeing. Um, and then, you know, a more prepared institution allows for,

13:06you know, more competitive pricing, which allows for more volume for our customers and

13:09better pricing for the American borrower. So it's really one of those win-win scenarios.

13:16I think what's so interesting about that is that, as we think about, um, you talk about a trusted

13:20source and where that information is coming from. So if someone just has a chat, GPT is like,

13:25hey, tell me about the mortgage market, whatever. Um, in this case, they know that it's locked data.

13:31They know where that data is coming from. They know that, um, the, that someone's looking at

13:36that and being like, is this correct? And I think that's, that's the further step because we can all

13:40ask chat GPT for stuff or cloud or whatever. It's like in the, the way that this is so helpful

13:46from my

13:46perspective is that what it's based on, like you, um, you know, as a customer, I don't have to worry

13:52about like, okay, where is this coming from? Is it really right? The source itself is what I'm

13:57trusting. Right. All right. And I don't know, um, how, how you interact with your chat GPT,

14:04but half the time I have to tell mine, like, don't placate my feelings here. You know, it's

14:07like, I don't want to guide it and get clouded by you reading me and trying to answer tailored to

14:12me.

14:12This, you're exactly right. Our virtual economists, there's, there's no sycophant nature there. It is

14:17going to be purely, purely facts based. And because, you know, we are very tight on, on the

14:22box that our LL, LLM lives in, um, we're able to be very, very, uh, disciplined with ensuring that

14:30the data that comes out of the economist is reliable, um, as, as accurate as, as can be.

14:36And, and, you know, trustworthy for our customers, because we're, we're very discreet around what kind

14:41of data are we comfortable with the economist answering. Um, and a lot of data, all of that focused

14:47around, you know, rate market predictions. I think that's so important. We so see so much

14:53misinformation, but now it's not like misinformation on a headline that someone's reading. It's like,

14:58they don't even know it is. Cause it's coming up on, on what they've asked their AI assistant.

15:02And they're like, Oh, okay. I, but I think the, the key here is like, you guys are focused on

15:08those lenders as they're making decisions. What is the information they need to know

15:13so that they can, you know, run their business better. So from my perspective,

15:16that's just so much better and, and so much more helpful.

15:21Thank you. Yes. Hear, hear.

15:25Okay. Well, let's talk about the third annual summit coming up. Right. Um, so I've been lucky enough to

15:30attend, um, both of your previous summits and, uh, be on stage. I'll be on stage again, um, this year

15:36coming up very excited. So tell us a little bit about what to expect and, uh, why listeners should,

15:42should attend. Yes. Well, first of all, we're very, very grateful to have you on stage again,

15:48Sarah. So thank you for coming. Um, we are extremely excited to kick off our third annual

15:53summit as you, as you mentioned in Scottsdale and the first week of February, February 1st through the

15:583rd, uh, people can expect we're going to have, you know, over 500, uh, attendees across the mortgage

16:05and capital markets ecosystem. So our customers are third party partners on our platform, lots of

16:12industry personnel to be able to come and speak about different topics that they have significant

16:18subject matter expertise in. There's going to be multiple tracks for people to, to attend as well.

16:24Tracks focused on AI and automation, fat tracks focused on, uh, the Indian capital markets ecosystem,

16:31um, facts tracks on focusing on the origination side and growth in the business. And of course,

16:39uh, one year and dear to my heart on the product side is our, our tech trendsetters, which is really

16:44our product showcase track. Um, so lots of content, uh, lots of different speakers. You'll hear from

16:51customers. You'll hear from, um, industry leaders. You'll hear from optimal blue employees, uh, all throughout

16:57all the four of those tracks. And then of course we will have our general session on stage, which is

17:02where we'll unveil some of our large initiatives that we've been working on undercover behind the

17:07scenes that we're really excited to share with our customers, uh, on stage following Joe Tyrell's

17:12opening remarks. So you've seen growth in this summit every year. Why do you think, you know,

17:18what is it that people are getting out of it that, that makes more people come every year?

17:22I think a lot of it was, uh, making sure that people were hearing from others, not just optimal

17:29blue employees, making sure, you know, it's not just us kind of spewing our sermon out there for

17:34everyone to, to hear, but being able to bring together, you know, the true network effect that

17:40optimal blue, uh, brings to the marketplace, having that show up in person, live and in person at the

17:45summit. So hearing from leaders of different lenders that specialize in non QM, for example,

17:51and some of their experiences, we see those products on the rise, um, as well as, you know,

17:56folks like housing wire or the MBA coming in and being able to talk about market trends and what

18:01they're hearing, um, as well as just continuing to be that, uh, that central hub for folks to engage

18:09other partners where they want to see their pricing return, for example. Um, and making sure that we kind

18:14of cover that full spectrum of, you know, tech stack needs for our customers, all centralized around the

18:20capital markets theme. And I think that's been very important in some of these more challenging

18:25rate markets to kind of hear how people are getting creative to navigate these, these waters as we

18:30continue to kind of see, uh, see rates stay where they're at, um, and make sure that, you know,

18:36they're keeping up with kind of the latest and greatest on some of these creative interpretations

18:39of way to grow the business, ways to manage your margin, all of that wrapped around how optimal

18:44blue is building products to support, uh, all of those new ways to interact with our current

18:50mortgage market here in 2026, 2027. Well, and I think the timing of that one's going to be really

18:56great. I mean, we know we're in a, we're in a difficult cycle in some ways. Um, but also from

19:02my perspective, that's when people need information the most, that's when they need their tech partners

19:07the most. So I think it's a great timing from that perspective. Yes. Well, Erin,

19:14thank you so much for being on always a pleasure to talk to you. Um, looking forward to, uh, seeing

19:19what you guys are doing the rest of this year. And of course, being at your summit.

19:23Can't wait to see you. Thank you so much for having me, Sarah.

Optimal Blue’s Erin Wester on using AI to meet the challenges of this market

Category

Transcript

Comments

Recommended