[David] Allen Pollack is here, our Mr. Tech, with a Tech Update. Good to have you here, Allen.
[Allen] Good to be here. So is that Mister with an R or is it M-I-S-T-E-R?
[David] Whatever’s the most, honoring way, Mister for sure.
[Allen] So a couple all kinds of great stuff. I hope everyone got to watch the World Cup game yesterday. It was what a game. Yeah. As many of them were but yesterday’s game and the finish was just something crazy and to just even watch Messi cry at the end after they got the runner-up awards and on the field, it’s just, takes your mind off a mortgage for 20 minutes. Or we should say 100-plus minutes since they played a long game yesterday, but it was great. And you know what’s funny, is they brought Ted Lasso out, who introduced- Yeah … Justin Bieber, which was pretty funny. And whether you like the music that was at the World Cup or not, the halftime show blew away 10 times. And people I was with were all saying it. I was at a party , the halftime Super Bowl show. Yeah … it was interest- interesting to see. Yeah. But let’s talk about the tech world. I didn’t do any research on what kind of tech was used for the tournament and all the games. I know all kinds of crazy tech was used. But I will tell you something very interesting. When you see football games, especially near the end even when a game… if it’s a division game, there’s always people leaving. And I’m from the New York, New Jersey area. I grew up going to MetLife Stadium. We used to call it Giants Stadium, all that great stuff. But the aerial view even 30 minutes into the ceremony, I was blown away, unless the aerial view was from prior, which I don’t think it was. There was barely anybody leaving. That place was just packed shoulder to shoulder for everything. It’s a different type of fan, and I think, I feel like this year this four-year cycle, that more people have gotten into soccer than ever before. I don’t know if everyone feels the same, but it’s been interesting.
[David] Oh, I agree. Yeah. There’s something magical about this year. Fútbol. Yes.
[Alice] Yeah. Fútbol. Yeah. Yeah, I would agree. Great comments. Thanks, Allen. Yeah. I think so. Yeah. Yeah.
[Bill] Yeah. And, yeah. And, Alan, totally agree. One of the things that’s fascinating i- is listening to soccer football purists debating whether Saturday’s third place game was a better game than the finals, right? Yeah. Saturday’s game and if… For anybody that didn’t watch it, it would almost be worth watching parts of it just to hear the announcers, who were… they were having the time of their life. Oh, isn’t that true? Yeah. And I equated it to an all-star game where, defense was not the strong suit, let’s say. But from a soccer purist point of view It wasn’t as technically a good a game, but it was really entertaining to watch- Yeah … versus a one to nothing game on Sunday that, in its own right, was a fascinating game to watch based on the, obviously, the skill of, all the players- The players out there.
[Allen] Yeah. And you know what’s insane? I know, the NFL and other MLB, e- everything we have has young folks. there were 19-year-olds playing in the Cup, which was just insane.
[David] That’s insane. Yeah. Some talent. Yeah. It’s not about age- Well- … it’s about talent. That’s amazing.
[Bill] The, top driver in Formula 1 auto racing- Yeah is now 19, and when he was racing last year, he didn’t have a driver’s license for the first half of the season. And by the way, he failed his driving test the first time because he rolled through a stop sign. Yeah, that’s great. And his response was nobody’s ever really taught me how to stop a car.” Yeah, you don’t need it.
[Allen] That’s great. That’s hilarious. Yeah. That’s very hilarious. All right. Let’s let’s talk about some tech stuff. I got some pretty cool things going on in the news. By the way I did bring back a joke it was written by AI. I didn’t like any of them, but the final one in the end that I thought was the best was “I asked my AI to help me lose weight, so it deleted my Uber Eats app on my phone. Problem solved.” Yep. Yeah. All right. There is a site out there it’s called Hugging Face, which is an AI… It’s a site where people share code and AI code. AI got hacked by AI, and then it called for help, and then it got blocked by AI. So check this out. It broke this morning, July sixteenth, so last week. I think some of the news was updated today. Hugging Face, one of the biggest AI platforms in the world, got hacked last week by an autonomous AI agent. The attacker’s AI ran loose inside their systems for an entire weekend, made seventeen thousand recorded moves, stole credentials, and spread across multiple servers. Hugging Face, the security team at Hugging Face, did what anyone would do. They tried to use AI to investigate the attack. The AI safety guardrails refused to help them. It was blocked. Because the attack logs contained real exploit code, the AI couldn’t tell the difference between the victim and the attacker. And so in the end, they wound up having to use human intervention in order to further investigation and fix this thing. So- Wow. For all the mortgage companies that are relying on AI to figure things out, you know, just remember AI does not solve all of the world problems because they couldn’t solve these problems. Yeah, that’s a good- Now- That’s a good, that’s a good point. This one’s important, more in Alice’s department. It’s a crossover between compliance legislation and technology, which is that Fannie Mae’s AI mandate goes live August sixth. That’s not too far away, folks. That’s two weeks away. Fannie Mae’s lender letter, it’s LL-two-oh-six-dash-oh-four on governance, takes effect on the sixth. That’s seventeen days from today, actually, so a little over two weeks. Their AI and ML governance framework has already been live since March third. And you have to keep in mind that there is a firm, there’s an int– there’s an a li– a legal firm that’s involved all the time, but that performed the industry’s first AI governance limited attestation, just released what’s called the AI video inventory letter, and it’s a ready-to-use package, including a draft vendor letter disclosure schedule, and six-category vendor map tracker, model risk tier, all this other funky stuff that you probably never heard of, folks. Anyways, you need to go get it. So I’m gonna say it again. The attorney is James Brody of Brody Gap, and it’s called the AI Vendor Inventory Letter. And it’s a ready-to-use package that you can work with. So pay attention. Very important. Remember, you need to understand what data goes in the AI engine, where the data goes, who touches it, who uses it. Oh, yeah. It’s critical because you are responsible. I can give you one quick tip on this, that folks that implement AI, when you talk to your vendors, they can be very successful at leveraging AI bots and technology with your data when they use aggregates of your data. In other words, their system or your system can call out to the AI with data bits or aggregates of the data, and then do the processing of the actual PII data on your side. So meaning you can’t hand everything off to an AI engine, and you need to verify that. If you need consultants, there’s lots of them out there. If you’re in the middle of an AI project, you wanna make sure that you pause that final go live or that integration or contract send until you’ve done this. It’s not to be scared. It’s not to push you off. It’s just to preserve the quality- Yeah … of where your data’s going and the security of you and your borrowers. So keep that in mind. And in that sense- before I move on to anything else, David, any questions on that?
[David] Yeah. As a matter… I wanna sa- give you a shout-out for your services because you’re one of the companies that can come in and help them. It can… Are you not?
[Allen] Absolutely. I definitely can.
[David] Okay. Just wanna make sure- Thank you … that was well understood by our listeners. Yeah. Meaning, help.
[Allen] Feel free to reach out to myself or David anytime. I’m always available. So let’s move to the next one, which is this is interesting. Rob Prisman, I love his newsletter and many other ways to get data in the industry, but he just does, he has from the… I’ve been in the industry the least amount of everyone on this podcast and I’ve been over 25 years now or in that range. So I’ve been reading his stuff since that time as many people have been. But it’s interesting, he wrote this weekend “AI governance is now a competitive advantage.” And what he basically said is his thesis this week, the question of whether AI can do mortgage tasks. Can it classify docs, automate workflows, support underwriting? And he’s saying it’s largely been answered. The new question is whether you can explain and defend what your AI did. He said, “Regulators, investors, auditors, and consumers now require not just with confidence that a decision was made correctly, but a clear understanding of how it was reached.” That’s the most- Yes … important part. And I’m not gonna read through everything. You should go read it yourself. But I will tell you the key warning that Rob said is, “Organizations that focus exclusively on automation risk overlooking the more durable value in governance infrastructure.” Model oversight, data stewardship, and decision transparency. So what he’s basically saying is be very careful. Your strategic advantage could be that you are able to explain and understand and not rely completely on it. So no automation risk overlooking the more durable components. All right. With that being said, let’s talk about Dark Matter. So their AIVA, that’s A-I-V-A, has processed now over five hundred million documents and is adding LO command centers to their technology. So their intelligence layer, which is called AIVA, has embedded inside of the Empower LOS, and it has processed more than five hundred million documents. And what they say, it’s a shift as moving origination goes from doing to reviewing. The platform handles the repetitive loop, people handle decisions. So hats off to those folks. And then David, this one’s critical and I’m gonna tee it off to my own company but let me just bring this up. It’s from Retr. Everyone knows who they are, R-E-T-R. They’ve got all kinds of amazing analytics. So they’re saying, “IMBs are losing the most ground.” And it says in their 2025 full year borrower retention analysis across 1,100 mortgage companies. Now folks, that’s not 1,100 originators, it’s 1,100 mortgage companies. Industry average retention has declined from 40.6, 40.6% in 2024 to only 38.6% in 2025. That’s significant. I just… Y- you all understand what I just said, right? Our retention, the industry average retention, It wasn’t even at 50%. We’re talking about our retention just declined from 40.6, which is dismal, to 38.6%. The divide is widening sharply by type. Banks and credit unions, so financial institutions, folks, they were relatively steady, 48.5%. How about that? Financial institutions are holding stronger retention than the rest of the industry. However, that’s still not great. But remember, financial institutions have a larger portfolio of different- Yeah … types of products and services. They’re trying to get more into advisement services. But in addition they spend a, a little bit more money. I’m not saying that they spend more than you spend, but they spend more money on the lifetime of the customer. I think it’s about a seven-year total investment the way they look at it. But independent mortgage companies, so IMBs, this is gonna hurt folks. IMBs went from 29.6% to 26.9%. Wow. That’s horrible. And mortgage brokers dropped from 30% to 28.4%. So you can get this, it’s on Google. But we need to do a better job, and technology is the answer, but the wrong technology is the wrong answer. And I’ll tell you this Technology totally is a wrong answer. This is a game of relationship, and how you do the best job of retaining people is through the people to people- That’s right And so two last pieces here, David, and the last one because my company plays heavily in this area, and we’re new and I’ll do a, a shout-out to myself. But let me tell you first, the gentleman, I don’t know this person, but I saw them a couple weeks ago on LinkedIn, and I saved this, and it fit appropriate for today. It’s a gentleman named Andrew Adams, so shout out to Andrew. Middlesex and Essex, MA. He said, “Cheap and easy to build is the acquisition strategy, not the business model. The business model comes later once rebuilding somewhere else costs more than just paying the new price. And if your whole stack was built by prompting an AI without anyone actually understanding the code, you have no idea what it would take to move it. That’s not the tech problem, that’s a leverage problem. The tool companies know it too. The tool companies know it too. What happens to your operation if the per-seat cost goes up four times?” So what he’s basically saying is you have to be very careful. Everybody out there can build technology, but is it secure? Can it scale? If you’re not an engineer building it, but you’re prompting it because you have a great idea, that doesn’t mean that’s production-ready or a world-class secure and compliant platform. So you gotta be very careful about what you build and release and what you actually use out there. And with those two things said just a, a shameless plug to myself, David. With what they said about IMBs dropping under twenty-seven percent, I have a company I spent a year and a half building, and we’ve just recently launched it, and we’ve got our first couple hundred users we’re putting on it right now. It goes beyond just mortgage. We also have a tour company that does boat tours that we’re putting on it. But it’s not that loan officers don’t care about that twenty-seven percent though, folks. It’s at every touch point after an application feels automated, and borrowers can feel it. Meaning they feel like they’re just getting automated system messages over and over again. They’re lost in their inboxes. They’re just completely automated. That’s what my company solves. My company’s name is Sendzie, and it’s a customer experience platform that makes every interaction, whether it’s a welcome, a referral welcome, a pre-approval, closing, loan anniversary, it doesn’t matter, it feels personal and made directly for that person, and it’s not a CRM. It’s not a blast email platform. It’s an experience platform. So if you wanna know more, you can just reach out to me. And if you’re interested in anything else I talked about today, you also can connect with me. It’s Allen, A-L-L-E-N, @tms-advisors.com.
[David] That’s a good job on that comp… sendzie is… Could you just spell that for us, make sure we’re getting that right?
[Allen] Yeah, it’s Sendzie. S-E-N-D-Z-I-E. Z-I-E. And we… Yep, and we are– we’re working with with a number of mortgage companies already. Awesome. Congratulations on another new successful venture. You’ve done so many already. Yeah, thank you. Alan that’s cool. Very good. We’re gonna ride this one into the sunset one day.
[David] Okay. You’re not old enough to go into the sunset. You got too much to contribute. Keep contributing. All right. Thank you.
Allen Pollack
, Chief Operating Officer, Tech Consultant
Allen Pollack, a Mortgage & Financial Services Technology Advisor, is a subject matter expert in the mortgage origination process along with software product management and software development.
In today’s financial services push to all things Digital, Allen has been helping lenders and financial services solution providers align their digital transformation and technology strategies by removing the human element of risk, and automating processes that drive efficiencies and margins into profits.
Over the course of his career, Allen has co-created and developed technology business models that have birthed highly successful, innovative solutions and companies.
Allen co-founded and served as CTO of New York Loan Exchange (NYLX), a loan product eligibility and pricing engine (PPE) that made an immediate impact on the industry, scaling the company quickly and forming partnerships with multiple mortgage and financial lending companies. In 2012, Allen was a co-founder of a merger between NYLX and Aklero Risk Analytics that created LoanLogics, A Mortgage Loan Quality and Performance Analytics company. Allen served as CTO where he continued to bring new and innovative product solutions to the market that made a significant impact to mortgage lenders that reduced risk, scaled business channels, and grew profits in a very competitive and highly regulated market.
Allen is also is mortgage and finance technology contributor on a weekly live industry podcast, Lykken on Lending, and is launching a new podcast soon to be released, TechStack Radio, dedicated to technology and innovation in Financial Services.