Waiting for Rates to Save You? KP Says Mortgage’s AI Race Has Already Started

Speaker A: Ladies and gentlemen, welcome to a new episode of the FinTech Hunting podcast. We've got a very special guest for you today. He's one of the most respected voices in mortgage A, uh, market commentator, a technology forward leader, and someone who has a unique ability to connect macroeconomics, production operations, AI and loan officer reality all in one. Doesn't need an introduction, but I'm giving it to him. Please help me welcome back to the show Kevin Perennial, Chief Lending Officer and partner at PRMG kp. Welcome back, my friend.

Speaker B: Thank you, brother. I appreciate it. I feel like the, all the crowd that's gathering outside the USA Soccer game that's kicking off here in an hour and 15 minutes. They were just going wild for your intro.

Speaker A: It's because you're the guest. So they were all excited to have you on.

Speaker B: Well, I'm a legend in my own mind.

Speaker A: You joining me. We saw each other at ICE for a few minutes and said, hey, we got to jump back on. So I'm glad we finally were able to get this scheduled. I mentioned in the intro you're known for bringing macroeconomics and production and technology and all of that together. Where do you see the market right now and where is the industry pretending? What are they pretending not to see? What are. We kind of just have blinders on right now with the market?

Speaker B: Well, I'm going to break the seal early. AI. It's all the talk and um, it's, you know, there's been a lot of great technology partners and vendors. You know, you're connected to many. And um, they've been building and um, you know, last year was a year of ROI on AI spend. Um, I've seen a lot of it, uh, being done. We have a partner, um, we've been with for um, over six years in capacity. We've had an answer engine. Remember they used to be actually called Jane AI.

Speaker A: Right, uh, right.

Speaker B: You know, and then they rebranded when they had their series, uh, B round. And then, um, they're rock and rolling. $100 million in annual recurring revenue in an AI company. It's, it's, it's fantastic. And so, um, I guess it was funny because I did a, uh, with Mortgage, uh, with Mortgage News. I did a deal with uh, Zach, Sasha and Greg, um, Sasha Ster and Greg Holmes, uh, yesterday. And one of the questions was, was actually pretty amazing. Someone on there said, hey, how would I approach my leadership team? Who said, we're not going to invest any technology. Just, you know, you and the people just need to make the most of what we got. I was like, you know, that was an eyebrow razor. Um, but I get that because there's been a lot of technology spend, whether AI or not spend on people. Um, parts this year have not been as fruitful. Um, and, uh, companies want to make money with what they've already spent. So maybe they're, um, you know, just saying, we're not going to dig in yet. And the good news, that approach, by the way, is if you're a company listening and you haven't made AI investments, by the time you do, it'll be the best version of it it's ever been, and you may leapfrog your competition then. So, um, it's, it's still, uh, third inning is what I would say in the mortgage space. And we're probably going to play, um, extra innings in this game. So we have a ways to go. And uh, but you can start today or you could start later this year or, you know, you know, maybe, maybe some people need, um, rates to tick back down and give us a little, little revenue boost and get a little extra shekels in there and then, uh, you know, then make the investments. The second half of the year is traditionally slower for purchase business. It's still steady. Each of the last two second half of the year, you know, we were given a nice refi boost. Uh, we may not get that one this year. Um, it's. It's hard to tell. All the last couple weeks were good. Our locks were up in June, uh, fundings were up in June. So July funding should be up over June. And um, you know, this week's kind of a light, weird trading week and quarterly end rebalancing. So, uh, bonds didn't exactly do well, uh, today, here, July 1st and yesterday. But, um, you know, these things, they move. They're very volatile. So, uh, you know, I'm, I'm excited about the second half of the year. Maybe not everybody is. Uh, uh, but, um, I think we're, we're positioned pretty well and I'm excited to spread the good word here on this show with you.

Speaker A: Well, uh, you bring up so many good points. I'm going to dive back into AI and tech in a few minutes. But you mentioned, you know, rate relief. A lot of lenders are waiting for rate relief. Is that a strategy or is that hope or is that denial? Where do you see that?

Speaker B: I think the right strategy is continue to build out your education and product mix for non QM and equity lending. Those two products are a, um, solid base that are Way less rate sensitive and um, there's just a ton of it out there and I think we'll actually see um, a uh, resurgence of not that non QM has fallen off but I think there'll be a better second half of the year than most people think because the, a paper rate environment isn't you know, giving um, um, you know, consistently. Um, so that would be the strategy I would say. And then if you get a little boomlet, a micro burst, whatever, um, you know, uh, take it, take it when you can, you know, on, on the refinances. Of course equity lending um, is a little bit rate sensitive, definitely more than non qm and that could see a little boost. Um, we just rolled out yet another I think our third in house equity product and we're looking at some other ones. And so I, I, I think those are, that's probably the right strategy for the old school, um, you know, thought when it comes to dealing with just general business and rate flow and um, and of course you know there, there are many other strategies hiring in M and A. We're seeing a lot of M and A activity, consolidation, um, of course tech investment.

Speaker A: Let's talk about when you and I bumped into each other. We talked briefly about trigger leads. Right. So all the hype, a lot of the dust has settled but now lenders are faced with the reality of okay, where am I getting leads, how am I going to grow my business? Talk to me a little bit about that impact and really more what's kind of that go forward strategy now for people.

Speaker B: Well I think you know, for real lenders, you know, that are enterprise, you know, we're very concerned about TCPA compliance so the Privacy Act. Right. So we want to make sure um, that whatever lead we get, um, we have a proper opt in. You know, it's well documented but there are a lot of little companies that don't care and um, those are the ones that mostly took advantage of trigger leads. And so at the bureau level, um, the bureaus get hit um, on the trigger lead legislation. And so um, I, I think generally that practice is, is, is pretty much gone. I mean there's still trigger leads, um, but I would say by and large it's, it's significantly reduced. And so um, you know, where people get leads is back to old fashioned, you know, ground and pound blocking and tackling, you know, uh, social media strategy, uh, you know, meeting with real estate agents, um, trying to find some alliances and some niches here and there. And, and I think that really helps, um, A lot of the independent mortgage bankers, um, and of course brokers at the local level who have good marketing as well that are not trigger leads. Um, I think it really helps them cultivate that purchase business which is a setup for refinances later. So I think it's a little bit back to basics and um, it does make it harder for some consumer direct companies to not have the triggers unless of course they own the servicing. Um, and that's a huge advantage. So that battle we've talked about in the past um, between originator and servicer, I think that will continue to roar, um, but not as much in a high rate environment. Um, and again I think where a lot of technology is helping everyone, is helping local originators and um, independent mortgage bankers become consumer direct desk. The technology is helping people fight that guerrilla tactic warfare and then also they're local. And so I think that is really going um, to swing maybe that servicer originator battle.

Speaker A: I love that point you made because people always think, okay, it's not going to help the local guy. The technology is just the big guy who's got all of it. But to your point, if they can leverage the technology and they still know the local area and the nuances of that market and have established relationships, that can truly be a win win. Uh, right there. You mentioned AI earlier. We'd be remiss. We're uh, on a fintech hunting podcast. Let's talk a little bit about what's working in AI. Where are you seeing? I know you mentioned capacity and you've been using that for years. What are some of the other tools that you see that they're getting an ROI right now, whether it's in lo agents, whether it's in loan predictability. Where are you seeing AI producing measurable roi?

Speaker B: Uh, I would just start by saying everywhere. And um, it really is a huge productivity boost. And so I think leadership teams have um, a conscious decision whether to use it to cut costs or to drive business. And um, we don't want to fire our PRMG family. We love our teammates and, and we are working our AI strategy and sales first and I think that's the right path, um, to do it. I think it says a lot about the uh, intentional leadership at companies. So um, we have an AI, um calling assistant that we've rolled out two months ago and um, had our second loan fund uh, this week. Uh, and yeah, so it's, it's making 2,500 calls a week on just 3,500 contacts. We have tens of Thousands of contacts. Of course, we're starting with our portfolio first because you have an existing business relationship and you're saving the consumer money a little bit easier to defend that in a regulatory manner. Uh, but we've had opt in language and consent language on loans, uh, since July of a year ago. So we have a, we have the rest of our portfolio that we could look at. We, um, which of course, even if it came from a tpo, uh, client, we can always, you know, call on that client on that bottom, the borrower and then get back to the TPO agent, which is something we're intent on. And then of course we can expand into um, anything not in the portfolio, um, depending on uh, the circumstance. Um, if we want, um, we, we haven't gone that far yet. We're just working the portfolio. Um, and what I'll say the reason why sales first is important is if you load up new opportunities and new business now the fulfillment operations teams are busy, so then you come and give the lift with AI behind that. So whether it's AI underwriting or document data extraction, things that, um, human out of the loop, um, inside processing and underwriting, um, you know, those are, those are things we want to do after we've loaded them up with more than they can handle. Because if you went there first, you wouldn't need some of them. And we don't, we don't want to do that. That's not our intention. And I think sales first is the right approach in implementing, um, AI and technology spend well.

Speaker A: And I love how you mentioned it starts with leadership intent, right. And really understanding are we just trying to be more efficient and cut everything that we can, or are we really trying to, uh. And so when you take that approach of we're going to focus on growth first, we're going to focus on sales, we're going to do what we need to do there. Then you've got the rest of the volume that people are saying, hey, instead of us having to throw human spackle at it because rates go down and we have so many more deals, you can just be more profitable at that stage and keep the team that you've already assembled. When we talk AI, there's a lot of discussion of do you go with niche products? Do you try to go with an enterprise platform? Is it really more of a hybrid of balancing the two? Give me a little bit of your insights on that.

Speaker B: Well, I think it depends on like at your company, are you a build versus a buy shop? And I think most of our audience is a buy. But what's interesting about um, the enablement of AI so like we have um, you know uh, anthropic cloud workspaces and we also have OpenAI chat workspaces that is enabling us to be more of a builder and we are building a lot of new stuff that is accretive to our entire ecosystem. We're on a compass and um, they've done a great job of becoming more web, uh, based and opening up their API architecture for us to kind of customize our stack. And so um, I like that ability for us to plug in what we see is the next best thing. So for example we have lone express originator uh through lotusoft which is out in front of Encompass. So all of our origination channels um, are uh, have been moved over or in the process. Our non del core channel just finished up yesterday. So a year ago we did wholesale retail is harder and more robust. We're in the process of that. We're about 30 done. We should be done this summer. And then our delegated channel is already on it too. So, so we have this web based gate out front that is very API, um, and AI friendly in addition to the data side and of course the underwriting piece and the fulfillment piece in the back office with the compass. And we think it's a good powerful one two punch um, without having to make an LOS switch. And so uh, we're big fans of that approach. That's been, that's been a big thing that we've done. Um, I think there's a lot of Encompass users that might want to take that approach and enlisting uh, to you know, to that that's been by far our biggest project. We've been working on um, for a year.

Speaker A: So with that I, I'm really hearing that how you vet technology providers and service providers is starting to change from the perspective of I don't necessarily need this one to do everything but if I have open AIs and I have the connectability that I have then with some of the tools maybe I can build some. Whether you're using um, a cursor or uh, any of the other tools to build some of it, you can build some of the tools and you can also extend functionality that maybe that provider never even thought of creating because you have the open architecture and everything like that. So do you vet differently because of that?

Speaker B: Yeah, definitely. We want to make sure that um, an AI partner has interoperability amongst different AI models. Did they write their code to maybe use an anthropic model or Gemini model or some other model, um, whatever they see, they see fit. Those, those individual AI tech vendor partners, they're really experts in one spot. But I will say that generally, especially if you're a buy shop, you know, as a lender, you are always looking for a potential vendor to maybe take out two to three others and consolidate with all the same functionality across. Right? So um, you know, that's always something in the back of our mind, um, you know, trying to make sure that we, you know, we don't have, you know, a bazillion different vendors. I remember this uh, exact moment four years ago when the industry really dropped like a rock in July of 22 was brutal. We had 252 vendors and we went through every single one of them like we don't need that one, going to cancel that one. This one, we need this one will cut costs and we ranked them all. And so uh, it was something, it's now an exercise. We do, you know, all the time now and um, minimum quarterly. So you know, I do, I do think that it's, it's, it's like initiative specific and what your priority is at a company and um, then also thinking are you build versus buy and if you're more buy, can you buy something that maybe consolidates a couple different ones? And so I think that's probably the logic that most executives are thinking right now.

Speaker A: You always have so many great insights. As we talk tech and we talk AI, there's a ton of new AI solutions that have come on the market in the last 12 months. How do you start kind of reviewing, vetting, looking at, is that something you want to pursue? Is it, hey, that's kind of nice, but we've already got something that we're kind of leading down that road. How do you kind of weed out what's the real, who are the real players who aren't. Because everyone thinks they're going to come in, they're going to disrupt the industry, They've got the coolest AI solution in the world. But how do you figure out what should plugged in and what is not worth your time?

Speaker B: I think if something is one of one, then you could make a pretty confident decision even if you're an early adopter or fast follower. So that, that, that is, you know, are there, is there a competitive moat with a vendor? If it's something that's becoming a little bit, you know, ubiquitous and you're see like, oh man, I've seen like a dozen of these things, then you can afford to wait, right? And then you can watch them iterate. Did they get better? Did they get customers that you could ask as a referral? Did they get money? Did they get more money? How did they spend their money? Are they good now that they have customers? That customer experience and success and support? Um, in addition to developing and reiterating, are they on old tech? You know, um, are they using agentic AI versus, you know, they built their stack a year ago on large language models and now that's out, it needs to be agentic. And so there's, there's a lot of ways to look at that. And um. And so uh, again if you take your time in the vetting process, you know, depending on how much time you take, I mean, gosh, you know, three to six months is like, you know, multi, uh, generational leap forward for AI tech now. So you have a chance to see that and see how that like oh man, last time I had a demo you didn't even do this, this and this. So there, you know, it's good it. And there is some patience, um, that would pay off there as well.

Speaker A: What are some of the solutions you're most excited about? What are, what are some of the areas that you see AI making the biggest uplift right now?

Speaker B: Um, you know, obviously lotusoft is a big one for us. That's really helped us. That is a base layer architecture for our entire company. Uh, we put it together with um, um, a few other systems and uh, that are proprietary and so it's called Leo Lone Express Originator. So we have multiple um, systems talking to each other. So that, that's been good. Um, you know we're a total expert shop and um, we uh, we, we like uh, to align ourselves with um, a partner that's been developed in their full stack. So um, we like Sella scla, which is their AI voice, uh, assistant. And then um, Cindy Cindie is a really fantastic human out of the loop, uh, texting platform. Um, and so we've been using um, all of those, um, those are the ones I like. Of course. Capacity. I mentioned that we've been on them and we deepening that relationship even six plus years in um. I really have my eyes on Grove's IQ from a processing standpoint which um, you know I met with um, Chris and Elena Groves when you were in Vegas in your um, studio up there at the Fountain Blue. And um, I was able to uh, to really work together with that group on their advisory board. Um, I really like what they're Building and they're doing it the right way. So um, those are probably some of the most exciting ones that um, I would stick my name to right now. And there's you know, probably 30 to 50 that I've done demos on that I'm keeping tabs on and when the time is right, um, you know we'll start. Oh of course, Friday Harbor. I failed to mention that we, we are um, we have great income calc with them. Um, Halcyon has done a fantastic job of gathering a tremendous amount of data. Um, you know the 4506 process is now dead. They killed it. It's 8821C data getting uh, you know, getting your data straight from the irs, getting good clean data accurately and quickly to be able to enable and enrich and enhance using AI too. So don't sleep on the data aspect either. Um, they've got true calc and they've got some great um, you uh, know, income stuff there too. So there's a lot of things that, that we really like and that we're using and um, keeping tabs on um, many, many others.

Speaker A: What's next for these next six months for the industry kind of where are we heading? What are some of the things that, that you're looking on out for?

Speaker B: I, I think a lot of, a lot of it depends on each lender individually. You know, um, I will say like retail, um, only independent mortgage bankers and distributor retail. Um, it's a little challenging when uh spring purchase season ends and so it's hard um to um, to, to drive in revenue when it's seasonally not there or maybe rates aren't necessarily as, you know, as kind. But I, I, you know I think um, there's still great business to go get. Um, we're a multi channel lender so you know we have, we're a, a diversified originator. We also service and so our book is there to help us out. I, I think there'll still be some consolidation this year um, if the rate environment doesn't do what it did each of the last two years in the second half. So be looking to see um, you know, um, I mean I would say from a recruiting standpoint, you know who's not liking where they're at and you know peak recruiting season. Model Match, we love Model Match. They just threw out all kinds of access to data which we can use, you know our cloud workspace and um, they're really good and helping us with um, you know, recruiting, finding um, real estate agents, um, loan Officers and also um, you know, borrowers. And so um, be interesting to, you know. Eric Levin doesn't like it when I say uh, uh, peak recruiting season. Um, first of all he doesn't like the word recruiting. It's talent acquisition or uh, but you know, I don't care what he thinks. There is a peak season and it's coming up between now and the end of the year. So I think there'll be a lot of movement um, out there because um, people, you know, the divergence in what a platform and a lender can offer to everyone that it serves from originator for us, originator first to the consumer, um, it is, there's some stark differences and I think people um, are wanting to see that. Um, you know, and so I think there'll be a lot of movement um, this year and you know, same as it ever was.

Speaker A: One last question. You and I could talk for hours and I greatly appreciate your time but you travel across the country, you attend all kinds of national shows. But what I have seen lately, and I wanted to kind of get your opinion on this is the state organizations have really started to step up from the quality of content, the quality of the speakers. I know you're heavily involved in the California mba. Tell me a little bit about what you're seeing as it relates to some of these state organizations and why these conferences are so important.

Speaker B: Well, you know I'm on the road and kind of my LinkedIn videos really started as like an on the scene reporter. Like hey, if you're not here, here's what's going on. I'm happy to share it with you and I think you know, building that collaborative collective, um, you know, environment, um, and I have a very small circle on LinkedIn. Um, you know I, I, it's people that I hand picked. I'm, that I'm troll for likes or engagement. Uh, for me I, I, I go for impact. My, my network on LinkedIn, um, impacts my business. So um, I only do two videos at you know, max a week. So I'm like, I'm clearly not trying for that. Um, but I like to help, I like to talk, I like to spur discussion and um, it's brought a lot of um, opportunities our way that are um, unique and impactful. So I think being on the road and continue to be visible, um, you know, on LinkedIn and social media is good. I'm on the executive board for California Mortgage Bankers Association. We have our Western secondary coming up in August. Um, we'll have uh, that, that's coming up. And of course, um, you know, PMG has our leadership conference for our retail channel in September. Uh, Digimo's coming out. You, uh, know this TMC's got an event, of course, NBA annuals in Chicago, so not too far from you. So I'm, I'm really excited about um, a lot of the things that are coming up here. And um, and of course, you know, I, I'll, I'll probably leave, uh, the California NBA on that Tuesday night because the next day on the 13th, your boy here is turning 50. So I'm gonna have to get home and.

Speaker A: Awesome.

Speaker B: Yep, yep. Thank you. It's uh, it's. I know, I know. I, I shaved and I'm covering my grays, but I'm gonna dye my hair tomorrow at 8:30 so you won't even know it.

Speaker A: Well, all I can say is welcome to the club and we'll leave it at that. Ap, you always share a wealth of knowledge and expertise. You're always so giving of your time. I can't thank you enough for being on this episode of the fintech hunting podcast.

Speaker B: Thank you, brother. Go usa.

Speaker A: Awesome. Uh,