Shifting Tides: Rates, Jobs, and AI Momentum | Optimal Insights | Aug. 25, 2025
In this week’s episode of Optimal Insights, Jim Glennon, Jeff McCarty, James Cahill, and Kevin Foley discuss the evolving interest rate environment, labor market complexities, and the accelerating adoption of AI in mortgage technology. They share insight into how these factors are shaping the industry and the broader economic landscape.
Key Topics Discussed:
- Federal Reserve Commentary: Analysis of Chairman Powell’s recent remarks and their implications for interest rate policy.
- Labor Market Dynamics: Unemployment trends, immigration’s impact on job creation, and the reliability of BLS data.
- Economic Indicators: OBMMI at 6.57, 10-year Treasury at 4.25%, and expectations for a Fed rate cut in September.
- CMBA Western Secondary Recap: Optimism in the market, rise of non-QM lending, consolidation trends, and margin optimization.
- HousingWire AI Summit Recap: Rapid AI development, ROI evaluation, compliance and security concerns, and sustainable productivity.
Tune in to gain valuable insights to help you stay ahead and maximize your profitability in the ever-evolving mortgage landscape.
Optimal Insights Team:
- Jim Glennon, Vice President of Hedging and Trading Client Services
- Jeff McCarty, Vice President of Hedging and Trading Product
- James Cahill, MSR Account Manager
- Kevin Foley, Director of Product Management
Optimal Blue Production Team:
- Executive Producer: Sara Holtz
- Producers: Matt Gilhooly & Hailey Røise
Commentary included in the podcast shall not be construed as, nor is Optimal Blue providing, any legal, trading, hedging, or financial advice.
Mentioned in this episode:
AI at Optimal Blue – Innovation That Delivers for Your Bottom Line
Is your technology vendor selling real solutions – or just the idea of them? Too often, companies pitch big returns, offer free trials, and promise seamless automation, only to reveal during implementation that the tech doesn’t actually exist. At Optimal Blue, we take a different approach. Our AI-powered automation is already live, already delivering, and already saving lenders real dollars. We show you exactly where the ROI comes from and how it reduces risk – no vague timelines, just proven technology that’s working in production today. AI at Optimal Blue isn’t just about delivering software; it’s about empowering lenders to make smarter decisions, operate more efficiently, and stay ahead of the curve. Built around real capital markets challenges, informed by direct engagement with thousands of clients, and focused on delivering measurable value at no additional cost to clients, while others talk about AI, we’re delivering it – Innovation That Delivers for Your Bottom Line.
Optimal Blue Summit 2026 | February 23-25 | Scottsdale, Arizona
In case you missed it – Optimal Blue has announced the dates for its 2026 Summit! Join us February 23 – 25 at Talking Stick Resort and Conference Center in Scottsdale, Arizona, for a client-exclusive event packed with strategic insight and innovation. Attendees will gain early access to Optimal Blue’s latest generative AI and automation tools, dive into expert-led sessions on pricing, compliance and more, and experience hands-on demos of new capabilities. Plus, it’s a chance to connect with capital markets pros, integration partners, and mortgage insiders—all while gaining actionable strategies to maximize profitability and sharpen your competitive edge. The Optimal Blue Summit: where proven mortgage expertise meets modern innovation to shape what's next. Early bird registration is now open at Summit.OptimalBlue.com – space is limited, so register today! News release: https://www2.optimalblue.com/optimal-blue-to-host-its-2026-summit-february-23-25-in-scottsdale Registration: Summit.OptimalBlue.com
Transcript
Welcome to Optimal Insights. I'm your host, Jim Glennon, Vice President of Hedging and Trading Client Services at Optimal Blue. Our clients and industry partners have long relied on Optimal Blue for trusted insights and commentary. And these podcasts are an evolution of our commitment to keeping the industry informed. Let's dive into today's episode.
Thanks everyone for being here. Today is Monday, August 25th. A lot to talk about today. In the interest of keeping everyone informed, of course, we will talk a lot of current events today. We're going to meet up with Jeff and James just to talk a little bit about what's going on in the market. A little bit of fireworks on Friday caused Powell's also just a general market update. And then after that, we will meet up with Kevin.
Jeff and I will talk with Kevin. Kevin and I were both at conferences, industry conferences a couple of weeks ago. So we want to talk through conference, which was the AI Summit hosted by HousingWire down in Dallas. And then I was at Western Secondary with many of you about in Terranea. So that was a California MBA conference. So lot of great topics discussed there. We'll talk about what the topics were, what the mood was, kind of what the trends were, anything that we learned and give you kind of the up-to-date on that.
Before we go there, just in the way of data, OBMMI 6.57 as of early this morning. So looking pretty good there, getting close to that six and a half handle. We're not seeing quite a spike in volume yet as a result, but I think the closer we get to 6%, we will start seeing 10 years at four and a quarter. So starting to see better spread between those two, between the OBMMI and the 10
Yeah, but no big spike in volume coming off just rates being quite a bit lower now than they were kind of mid last week. So we'll keep a closer eye on that. All right. Let's go talk with James and Jeff about generally what's going on in the market.
Jim (:All right, James and Jeff, welcome. Thanks for being here to chat a little bit about what's going on with rates and the economy and the Fed and just jobs and all of it. where to start? Maybe we start with the most recent, which should be Friday's comments by Chairman Powell. think, you know, I used to, in school, they would teach you, back then it was Alan Greenspan. They'd say if, you know, if,
Greenspan's are slightly different than the last time he spoke publicly. That means something, right? And I think that carried through with Bernanke and Yellen and now with Powell and the words were slightly different, right? Jeff, you made the comment. It feels like the pundits are saying he's suddenly gone dovish, but it seems maybe a little bit more like he's gone into the middle instead of being that hawk, right?
Jeff (:Yeah,
right, right. He definitely made some comments, like you said, a little bit more focusing more on the employment situation. But it wasn't like a in that regard. as you said, any slight tweak has everybody in a tizzy. Reading the tea leaves here is what everybody loves to do. you know, but.
Jim (:Mm-hmm.
Jeff (:It was his first chance, main chance after most recent announcement to react to then those jobs numbers, you know, and the revisions that came out a couple days later after the Fed announcement. you know, I'm not sure if there's anything too surprising when you look at it in that regard.
Jim (:Yeah, good point. I I mean, let's talk about those jobs numbers again. We've talked quite a bit on this cast, but they seem to be getting fully digested by the masses, right? We had poor number in terms of non-farm payrolls that came out for July and early August and then huge revisions for May and June, right? But there's a lot going on there. We still have an unemployment rate that's relatively low at 4.2.
right, unless it gets four and a half ish. Historically, the Fed does not need to act, right, or it's not kind of a compelling case. So, and James, you had some interesting comments earlier just on the dynamics, right, of this, like we've never seen quite a market like this, especially in terms of the interaction with like the White House policy, right? A big part of this jobs story is what's going on with immigration.
James Cahill (:Thank
Definitely. Well, you know, back in the 70s and 80s as the interest rates came up in an effort to fight inflation, there was there's a strict balance that has to be met there, right? It's inflation and unemployment and how can you fight it? And they chose to fight inflation at the cost of employment. It was definitely a difficult period, but they had the backing of the administration, the Fed did at the time. It's very clear that the administration has made its decision that unemployment is the number to target this time around.
And so with the pressure that's on the Fed, the numbers do encourage that we could look at cutting the interest rates, letting a little loosen into system, trying to get employment moving a little bit better, even if it's not bad, having it move a little bit better is what they It's hard to say that that would be the wrong decision. It's just, it's the decision of the time, that it's a hard balance and there's an outside force that
sees the way it wants to go.
Jim (:Yeah. I mean, the inflation picture still looks good, right? If maybe tariffs were a little more certain or maybe if the effective tariffs were more certain, we would have cut rates by now because even though we're a little above 2%, as you pointed out, Jeff, anytime there's a fight against inflation by the Fed, there's a tail on it, a really long tail, right? And we're probably towards the end of that tail right now.
Jeff (:Yeah. Yeah.
Again, like James had going back to the seventies and eighties when Volcker was fighting inflation, that was a very obvious time where there was a super long tail. so, you know, it got down to 4 % and then it took a long time to get down further below 4%. So, you know, think some of this is normal that the Fed is continuing to fight inflation, but still probably generally headed in the right direction, albeit slowly.
But yeah, with these external forces, I guess if you call both the employment side and the inflation side of tariffs, what are the effects of tariffs on inflation? Is it a one-time increase or is it gradual increase that's a true inflation trend?
still have to look through it in that lens. And then on the employment side, as you said, effects of immigration. So we saw these really low jobs numbers with the revisions, really low jobs numbers. But if you look through through the lens of immigration has fallen off a cliff. And a lot of in employment are coming from immigration.
you know, maybe those numbers kind of make sense. So it's, you know, we're getting ourselves worked up at all these things through the lens of quote normal environments, but know, we're obviously not in a normal environment with kind of the, I'm gonna use the word, policies of this administration.
Jim (:Yeah, I mean, the unemployment, just to underscore it again, there's so many dynamics to it. But if you were to look at the unemployment data purely on its own, you would say, looks fairly healthy. But then you see mass deportations, there's immigration, like you said, is down to zero. Even
workers in this country afraid to go to work in a lot of cities, right? So they're needing to be replaced. So the economy is not needing to create jobs to keep people employed. Like that's the reason why we're creating very little jobs. Yet the unemployment rate is staying low. Like that has to be a big reason that that's happening, right? So I think more of the articles you read about the job market being sicker than it looks are probably right on. Right? I think there's a lot more of that kind of rhetoric out there and it's likely.
Absolutely true. Folks that are getting out of college right now, for instance, are finding it very difficult to find work unless they're in healthcare, really. That's where the majority of the jobs came from in the July number.
James Cahill (:Yeah, I've seen an interesting stat that it's roughly 6 % of college age men graduating with a degree are currently unemployed. And that's not only the highest it's been potentially in US history, it also now matches with the rate of same age men who did not attend college. So getting the degree is no advantage, effectively, over not is what the jobs market is kind of
flagging with that did pick up, but the rest of the market's very flat. And to your point about, you know, is it maybe a little sicker than we know? I spoke about it previously, but the Bureau of Labor Statistics and their perhaps hampered ability to get the numbers that they want and actually see what's happening is definitely a factor at play here, right?
Unemployment rate is 4.2 % based on reporting by the Bureau of Labor Statistics, which is only as good as they can get. So there's definitely a margin of error there to be aware of.
Jim (:Right. And as we pointed out before, 72,000 jobs in an economy as large as the US is like nothing. We could see revision to the negative in the next print.
Jeff (:Yeah, was been reading a lot of, you there's been a lot of fed speakers lately, obviously a few interesting things that reminded me of something, you one of the, think it was Goldsby was mentioning, you know, you're kind of looking at these jobs numbers as plus or minus a hundred thousand jobs, which to your point, you know, 72,000 jobs, plus or minus, minus a hundred thousand is, is negative, right? So, we are picking apart.
some are very rough around the edges in a normal environment. then to James's point, where the BLS is potentially their data collecting mechanisms and reporting capabilities, even tougher to get a read. So you take all that in aggregate, again, going back to some of the fed speakers I've been reading, more and more of these fed speakers are talking about
they kind of just saying, I don't know. There's a lot of uncertainty about which direction they should be looking. Should they be looking closer at the jobs numbers? Should they be looking closer at the inflation numbers? And then within each of those numbers, you can tell a story in either direction more so than have been in a really long time. the Richmond president.
Jim (:Mm-hmm.
Jeff (:talking about three out of four times, it's pretty easy to figure out what the Fed needs to do. This is kind of one out of four times a lot of uncertainty and not sure about which direction the Fed should go.
Jim (:Yeah, it's like they can't use that data dependent excuse right now. you may, does the Fed need to read the tea leaves a little bit because the data is telling you one thing, but the average American is telling you a different story and you have to make decisions based on that or you will be quote too late, right? With making a move this time around and maybe we're at the end of the tail, maybe things balance out with tariffs and, are we late to rescuing the job market?
or that has a lag on it as well. What we don't need, I'm sure, is five, 6 % unemployment, at graduate unemployment number to go from six to 78 to 10 % and really cause a weird dislocation in the market. Or maybe this, again, is part the White House's plan. I mean, they're looking to transform, not just
global trade, but also the labor market. Right. mean, a big thing that happened last week, for instance, was the Intel investment, right? The U.S. invested in Intel and they've really, that was done for one reason. was national security was one piece of it, but also American jobs is another reason that I think the U.S. government will continue investing in private businesses is to bring more jobs here domestically, right? But they're not all going to be white collar jobs. A lot of them are going to be in the trades.
that's been a promise of the campaign, right, is to bring more manufacturing jobs, especially to the US. And that's not necessarily always a college educated person who's taking those jobs. So there's going to be a transformation of that market. It's not going to happen in six months, right? It's going to take the rest of presidency and likely beyond to make that complete change.
Jeff (:Mm-hmm.
James Cahill (:I a good pivot off what you're saying there with the administration taking, I believe it's a 10 % stake in Intel, is is kind of a good view GSEs and what might happen there. So the administration was saying this start of the sovereign wealth fund that Trump had signed maybe in early January with an executive order that they wanted to start. They want the government to have a stake in US companies.
encourage manufacturing and other jobs at home and just to help build wealth for the country. Part of that, as we are waiting for the GSEs to potentially come out of receivership is how is that going to look? How is it going to work? How will they maintain the full faith and backing of the US government behind their credit? And the answer might be as they come loose that the government would maintain 10, 20, 25 % of a stake in them and consider that part.
of the sovereign wealth fund. That's quite possibly what we're looking at there.
Jeff (:Yeah, that's a good point. mean, and it definitely, you if the government's going to do that in a company like Intel, right, a publicly traded company no government intervention up until now, then they'll definitely do it in something like the GSEs, right? So it definitely removes that concern about what does You know, it doesn't mean too much if they're doing that with a company like Intel, I guess,
Jim (:Yeah. mean, that was, you know, sometimes it's tougher to read nuances in, you know, maybe a tweet from the president, different than trying to read, you know, Powell's pre-selected comments. But there were comments when I think back a couple of months ago where they were talking about, Pulte and Trump were talking about privatization. And a couple of times they mentioned like a kind of quasi privatization or semi privatization. So it makes sense that in the back of their mind, they may have been thinking this sort of de facto
sovereign wealth fund that was created essentially, I mean, it's not official, by putting 10 billion ish into Intel, exists now that the American taxpayer owns part of Intel by way of the government. So if they do this, something similar or even bigger probably with Fannie and Freddie, it could help with some of the uncertainty around the guarantee, of the uncertainty about where all that money goes now that they come out of conservatorship. could be kind of a win-win, I suppose if done exactly the right way.
James Cahill (:And it does make sense, employee stock option plans, or there's a lot of unions that one of the things they try and do is they try and gain a large or a voting block share of the company that they are representing the unionized workers of. That way they have the ability to vote. They have the ability to say, know, hey, we're going to move this to Mexico. Like, no, we're voting against that. We want to keep it here. Having the US taxpayer own it, the government own it, it's not just the ability to regulate.
and passed law that would stop it, they have the ability to vote in shareholder meetings and who's the CEO, who's the board and try and sway some control there. So it's a different approach to how the government can levy power against these large companies.
Jim (:Right. That could end up being the case. I think with Intel it was non-voting shares, but that doesn't mean in the next case that it wouldn't be voting shares if they decide they want to maintain some control. Like in the case of the GSEs, I think that's just a different animal because it's been nationalized for so long. If I can use that word.
James Cahill (:⁓
Jeff (:Yeah,
look, regardless of if it's voting or non-voting chairs, I think it's to think the federal government would not exert some sort of control these companies. the last couple of minutes here, we can kind of tie it back to looking ahead a little bit and some hard numbers. Jim, had mentioned...
OBMMI at 6.57 lower, not like crazy lower, right? mean, you know, the 10 year drops, obviously we've got some volatility Friday, the 10 year drops, but you know, we're continuing to drop here, going into this week. uh, CME has a probability of a fed cut in only three weeks now, three weeks out from the September meeting at 86 % right now. Um, so.
Jim (:Yep, I hate this.
Jeff (:you know, pretty well priced in that that's going to happen. Even with that priced in again, you know, rates have dropped some, but mortgage rates are still six and a half and above. you said, volumes have not picked up, right? think we need to, you know, we probably need another, at least a quarter point drop in mortgage rates to see volumes, you know, really start to pick up, ⁓ even more would be my sense.
Jim (:Yeah.
That's what happened last year. Last year, this exact time almost, we started to approach 6 % and volume spiked 50, 60%. It was a wild time if everybody recalls, but it was a year ago and it would be nice to see obviously going into the holidays because the holidays are otherwise very brutal in terms of mortgage and just housing market volume, but we may actually see something. If we keep going and you wonder if borrowers are sitting on the sideline waiting.
and hearing that the fed's going to drop rates and thinking that's going to have a big impact. And it might, but it won't have the immediate impact because I think, as you said, it's all priced in unless the fed surprises us and cuts a point or cuts a half point even. the bets are on a quarter point cut here in three weeks, as you said.
Jeff (:James, anything else, kind of last quiet week of the summer, but we do have at least one interesting number to look at this week,
James Cahill (:Well,
definitely. was thinking, you know, tomorrow to Tuesday will be a consumer conference. I thought was always something to kind of keep your eye on, especially as we have these conversations about, you know, kind of the fogginess of the market? What's everything looking like? There's also, well, how does everyone else seem to feel? You know, maybe it's just us in our bubble, but I'm definitely interested to see on Thursday is GDP for quarter two. So I've kind of harped on this before, but
Jim (:That counts.
Mm-hmm.
James Cahill (:GDP in quarter one was negative and people really responded to that, but there was a lot of front loading to get ahead of the tariffs. The GDP number looks pretty solid for quarter two. It's about 3.1 % is the estimate, but that's almost the pendulum swinging the other way. Now we've seen ⁓ imports really drop off, which is gonna push up the GDP number. So I would say be a little careful when that number comes out. There's probably gonna be a reaction,
of laughing about it, but there will almost certainly be a tweet, you know, look, look how great the GDP is, look how good it is. But that number is at least a little inflated as it comes back from the one time this year.
Jim (:Yes.
Right. So don't be fooled by the GDP number. That's the message this week because it's kind of a one-time gain that's coming from the shift in the trade deficit. Could happen more as we continue to increase tariffs on countries, but it feels like the biggest shock was, like you said, the negative number last quarter and then potentially three handle this time around. Yeah.
James Cahill (:positive potentially.
Jim (:I think that just about covers it then, gang. Good discussion. Good things to keep in mind here. Y'all, mean, we got three more weeks till the Fed meeting and then, yeah, GDP this week, and then we'll keep you informed on kind of what the bets are looking like and how we feel about interest rates in the coming Fed meeting. Thanks, gentlemen.
Jeff (:Thanks.
James Cahill (:Thank
you for having us.
Jim (:All right, we have a pretty cool update segment for everybody today. and I were planes passing in the night, if you will, a couple of weeks ago. So Jeff and I thought it'd be good to sit down and have a recap on the conferences that we went to.
Jeff (:Yeah, some pretty big conferences. Jim, you're at, you know, kind of one of the biggest regional conferences there is the Western secondary. And then Kevin, you were at the, wires AI summit, which is, I think second year they've done that really popular one, kind of the brightest brains in the industry at that one. some pretty important conferences out there that you're both able to go to. we thought we'd start with Jim, you kind of give a quick recap of.
general trends across the industry that people were talking about last week.
Jim (:Yeah, as usual, Western really well attended. I think the numbers were better attendance than the previous year we were out at Terranea. In California, I'd say that the vibe was even better this year. Obviously, coming off of 22 and 23, it was a little bit rough, this year I think people were in pretty good spirits. It was interesting, less and less talk about traditional GSE space.
in terms of origination, even though that's still a vast majority of what's going on. talk about the GSEs themselves, you know, with all the talk that we've had on this podcast and you see everywhere about a possible exit from conservatorship. But I thought it'd be good just to go through just a couple things were buzzing. of these might be obvious, but I'll let you know kind of what people were saying. Non-QM was a big one. You know, was, you know, kind of felt more real.
than the past few years. The past few years felt like a little bit more talk, right? We're going to get into non-QM, we're going to start selling it or we're going to start buying it. But there was folks actually originating it. And then that shows up in the volume. We certainly see it in our volume. We're up about 50 % year over year in non-QM. mean, granted that the total is still less than 10%, but it's up from something like 5 % closer to almost 10 % now and climbing.
you know, towards the end of this year. So it's certainly something that we've been focused on across Optimal Blue, certainly on the desk. We've been learning a lot more about non-QM, even though it's not, it's definitely not in a spot where most banks are going to want to hedge non-QM necessarily. But we're certainly getting educated on it and using the ton of data we have across our systems to help clients figure out who's buying it, what the pricing looks like, the major buyers smell like.
in terms of the right partners to have.
Jeff (:Yeah, there's a,
yeah, there is a robust marker. I know there's a robust set of buyers out there again, you know, maybe not quite liquid enough and particularly, you know, not enough, maybe fundamental, understanding of exactly how those things are priced where you can hedge it at lower volumes, but still plenty of buyers out there where you should feel comfortable originating those on a best efforts basis. Make sure you can, you know, you have outlets. I think that that type of market is definitely out there.
at this point. ⁓
Jim (:I think
you're right. I mean, you're seeing the big aggregators get into it now, which I think was a big turning point. You've got a Marihom and Penny Mac buying this stuff, ⁓ Mr. Cooper. Those are entities that are going to be way less likely, I think, to turn around overnight or disappear overnight if there is a market disruption like we've seen periodically over the years where folks just get caught. Even in the world of best efforts, you just get caught with a buyer disappears or is no longer buying a product.
Jeff (:Yeah, and I'll give one quick plug because I think it's an evergreen episode. Just a reminder, a couple of weeks ago, Mike and Vimy and I did a non QM episode and Mike gave a really great overview of the non QM market and different types of products. So a replug that episode from a few weeks ago.
Jim (:⁓
Yes, definitely go back and check that one out. If you're at whether you're a novice or an expert in non-QM, I think there's some really good information there. It won't be the last time we cover that on this cast. mentioned fundamentals. It jogged my memory that some of the discussions were folks getting back to some fundamentals versus like the crisis mode that folks have been in the last few years, managing headcount and efficiency liquidity issues and
Volatility, it's more back to just like optimizing margin, optimizing your counterparties, right? Getting back to who should I really be selling to, whether it's in the GSC or the non QM space, what broker dealers should I be working with? Nailing down your hedge effectiveness. Integrations, like how do I get all my systems integrated to get more efficient? Like how do I gear up for hopefully this next level of growth, right? Which we see it, we're seeing slowly, right? are, it's sometimes it's easy to forget.
that we're seeing double digit increases year over year the last couple of years in volume. But it doesn't feel like it because we're not getting these huge booms, right? But we could. mean, some of what went on just last week with the Fed that we talked about, could be, we could get one of those fall booms. How do we gear up for that without adding headcount? consolidation, just ⁓ &A. There were some sessions on that and just our meetings I think consisted of some of that discussion.
blind optimism like we had in:three, would say 24, we were just kind of hoping rates would go lower in the next six months. Like we are seeing rates slightly lower, but seeing volume go up nonetheless, because purchase business is continuing to improve getting real creative on how to, how to capture it. And then, uh, lastly, my panel was, I was on an AI AI and automation basically how to, you know, how to get better using AI. I think it's sort of like non QM where.
Like last year was just a lot of talk. People were saying AI, people were kind of proposing philosophically how AI could improve our business, our industry, but it's actually happening now. You know, and we're walking before running, in my opinion. I think some of the lofty ideas are still just that. They're difficult to attain, but you are seeing folks use things in their own lives, you know, like co-pilot to help kind of manage their day and to...
be more efficient, we're also seeing, you know, Optimal Blue for sure. And other platforms around the industry using AI just to make things easier and simpler and to start learning over time make things better. So I think that's something I think it's a good time to segue. I really want to talk to Kevin. We have not met up since we were traveling, just really want to talk to Kevin to see how the AI summit went and
just what the topics were there and what the main takeaways were.
Kevin Foley (:Yeah. So, ⁓ I, went down to, Dallas, other week and, went to the housing wire AI summit. was really very well executed. was, very, very popular standing room only actually for a good, good chunk of the time I gave up my, my prize seat, in the morning, during the break. And then I wasn't able to get it back until after lunchtime. So, lot of folks there, I think housing wire does a really good job of getting some of the best.
You know, mines in the industry altogether, one lot of great speakers. So our Civer, our chief technology officer there. He had a really Talked about a little bit on LinkedIn. He went ways to evaluate your return on investments AI, which, which I thought was really good. And he also talked about just the learning
and excitement curve AI, has sort of repeated itself with different disruptive technologies. You sort of have this exuberance at the beginning. say, this is going to change everything. And then you kind of come back down to earth and you realize kind of where the limitations are with the new technology, but eventually sort of build your way up to this sustainable productivity.
you have. And I thought he captured that really well. I think certainly, my standpoint at Optimal Blue, think we're farther, much farther along that sustainable productivity curve and understanding AI tools and how to deploy them and how to build with them and how to understanding the use cases where are going to work the best. so do think it's for lenders, there's still a lot of lenders, I think, who are
just in the process of starting to get their feet wet and certainly encourage anybody who's trying to think about, where do I go? How do I start implementing this? Leverage some of the partners that you have in the industry. I Optimal Blue is a great partner in that regards and helping lenders understand where they can see some of those gains just since we've been through it already. We've come out on the other side and learned a lot through the process.
certainly there to help any lenders who are for One of the big takeaways that I would say is people are building stuff really fast is the big that I'm getting. heard from big names, UWM, Rocket, who are out there and they are
you know, talking about building new capabilities in a matter of days or weeks, and, you know, building new AI agents kind of from here's, here's our idea to actually deploy it out there in production. And I think know, very impressive, that people are moving that quickly, but it does sort of make sense, you know, when you think about it, and particularly, you know, my understanding of, of AI agents and, you know, how you kind of, ⁓ you know,
actually go through the process of building these things. a lot of ways, you're really just connecting plumbing of tools and data sets that you already have and feeding that into an AI agent, which at its core way to think about it is just the chat GBT assistance API. you've got to have got all of these building blocks in place and you're
your job as someone who's building the AI agent is really just stitching them together. So there really is an opportunity to move very quickly start seeing some real value understanding and kind of get up to speed with the learning curve that's there, which by the way, you can use AI if you're just starting to get your feet wet and you're like, I don't really know where to start. Well, you can ask chat GPT, hey, where can I?
know, start to think about deploying, you AI capabilities to help maximize my ROI. It's, the really the first disruptive technology that actually helps you understand how to use it. so yeah, overall, very well done conference, summit, know, a of, a of interesting speakers there and, there's definitely,
I think an appetite amongst a lot of folks who are getting farther along in that learning curve to move very quickly.
Jim (:Yeah, a lot of buzz out there right now for sure about all the cool things that AI is doing, right? We got into some of the boring stuff at the end because there's this, you you mentioned how quickly it's evolving and how quickly it's being created. So we had to have a couple of questions on our panel about like, that might sound dangerous to people and it probably is, right? There's risks to deploying some of these thinking machines if they're not properly, if you don't have good policy internally or especially if you build your own, these things can't, or they
compliance issues, right? That's like a whole web of stuff that to worry about. Anyway, did they talk much about that at at the housing wire deal?
Kevin Foley (:Yeah, yeah, I think that certainly came up, and that's actually something that Siever has talked about as You know, there's a few different ways when you're thinking about building agents actually ingest the data or the tools that it needs to be able to understand your business and be able to perform functions. A couple of those, so there's what's called RAG.
is, can kind of think of, you know, your, your data being kind of bolted on, to, to the AI model, but you can do it kind of entirely within your own ecosystem. And then there's something newer that's that has a lot of promise called MCP or, think machine context protocol, which allows you, it's, it's kind of like, a framework for creating more plug and play type agents where you can connect them to different tools that go over the internet and whatnot.
And there's certainly, I think, if you're going to be injecting tools outside of your ecosystem with something like an MCP, a lot of security you've got to figure out as part of that process. It's certainly very powerful, but it's also, it's new and it's one of those things where the security layer that's available there is going to evolve over time.
but yeah, that's certainly, certainly a concern, I think both from, from the security perspective, also from the compliance perspective. where, where's AI making decisions, within, know, a workflow or your organization, that sort of stuff gets really tricky. there's a lot of applications I think where, you know, you, you're really, you can set the AI, you know, whether it's agent or whether it's copilot or how, are you, how are you, you know, configure it.
it's not the one that's actually making the decisions instead. It's just serving that information up to a human. think that's sort of the best way, I think, to manage ⁓ the compliance risk of kind of the AI running ahead of what your organization's compliance standards are.
Jeff (:That's great. I think the third thing I'll point out, you mentioned security and compliance, and you kind of touched on it with some of the things Siever had brought up, but just something, Kevin, you and I live and breathe every day, right? Building software is you run into with new software development, right? reliability is critical, really have to think about how these things fully integrate into your entire ecosystem.
and think about how you're going to maintain these and upgrade these. So the very obvious example right now is ChatGPT 5 just came out. system, a lot of systems are built on top of Microsoft using ChatGPT for a lot of these tools. So making sure you have a robust process in place to move from a previous model to a newer upgraded model. A lot of ways, that looks
the same way you would upgrade We're just kind of going through, you a lot of people are going through the first rounds of doing those types of things, but you have to keep a critical eye on that. while some of these things being built quickly are new and flashy, you know, with some of these complex ecosystems that a lot of our systems are built on top of, you just have to keep it, continue to keep a critical eye on those types of things.
Kevin Foley (:yeah, for sure. I think, and I think that's a good point. You know, when you're, when you're building something, you, do need to think about the maintainability, you know, what upgrades, how, is it going to evolve over time? I think this, you know, the, the, the chat GPT, and then also when your organization is going to be ready to take some of those upgrades. know, chat GPT, five coming out in the last couple of weeks.
certainly has been sort of an interesting, ⁓ interesting story even Sam Altman has, has, who's the CEO of open AI has, has mentioned were some, mistakes that were made and maybe also expectations were, were pretty, pretty high amongst its user base. sometimes it's not.
doesn't always pay off to be part of that bleeding edge. think some of the early adopters for a chat GBT-5, there was sort of a clamoring of folks looking to go back to their 4.0 models that they were more comfortable with. definitely a tricky landscape to navigate because you don't necessarily want to be beta group on a new model. But also you do want to make sure that you're
you're staying current as some of these trends are evolving because things are changing very quickly.
Jim (:Yeah. And the curve is steepening, right? I think about it. You could think about it in terms of like the iPhone, like not many people had iPhone one, right? I think my first was four or five, but you could see it, but it escalated so quickly from there to compare the first few to what you have now. And considering it was really only, I don't know, 10 years ago, maybe 15 that people got their first iPhones. I think it's just going to get wild here in the next few years. Everybody's projections are different on how it evolves, but super exciting.
Jeff (:Yeah, we can talk about this all day. I think there's a few other topics I wanted to hit on that we didn't even get to. So I'm sure this will be an evolving conversation, but it was really fun to hear the feedback y'all are seeing in the industry. think it lines up really well with what we're doing at Optimal Blue to toot our own horn a little something we all should keep a close eye on in our ⁓ personal and professional lives, So.
Jim (:Yeah, big thanks to HousingWire and the California MBA as well. Just two institutions that are doing just a great job of gathering and disseminating content.
Cool. Good segment, gentlemen. Thanks so much.
Jeff (:Thanks.
Kevin Foley (:Awesome.
Thanks guys.
Jim (:Okay, let's wrap this thing up. Big thanks, James, Jeff, Kevin. Good conversations. learned a lot today, so appreciate you. that's it for today. Join us next week for another episode of Optimal Insights, where we'll continue to provide you with the latest market analysis and insights to help you stay ahead. Check out our full videos on YouTube. You can also find each episode on all major podcast platforms. Thanks again for tuning in to Optimal Insights.