[David] Let’s talk a little bit about technology, Allen. There’s a whole lot going on there. And then Mark, were to come to you at the end for one of Mark’s rants, but talking about technology, the projections that I’m reading as I look into 2026 is the landscape is gonna be, it’s already been changing, it’s been fast evolving, but I think we are greatest changes and challenges. The challenges with the changes that are gonna be realized in 2026. And I think the cost origination, we’re seeing significant drop Allen, in the number of companies that are getting their cost origination well under half of what
[Allen] I call Bologna,
[David] You do, you call in Bologna?
[Allen] I do. I call Bologna.
[David] Talk about it.
[Allen] Just like the Fed’s unemployment numbers. Look, at the end of the day, the cost of technology and to have a tech platform as a vendor is skyrocketing. It is not cheap. If you want to implement AI, it’s expensive. You can’t just go with the automated building platforms out there. If you’re gonna launch something particularly, or try and embed it with your system, you can’t get rid of your engineers because you have legacy tech. You have to charge your customers more. You’re paying vendors 30% of revenue or between 20% and 30%. Everybody wants a piece of everything and you have to charge your customers. And a lot of vendors have slowly moved over to charging their customers for part of the integration cost because they can’t lose the 30% to an LOS or something else that they’re paying. And that’s just a few. I probably could sit down and name 15 more. It is expensive to be a vendor and I was, it’s funny you went into it this way. What I was gonna say when you and David were just talking about, getting into the business and starting out as a broker and, there’s a lot of people who said, I have the best idea to build the best pricing engine, or I wanted the best CRM or they weren’t necessarily, it’s not that easy and they weren’t necessarily meant to run a tech company. And what I mean by that is there’s a lot that goes into it. There’s relationships, there’s understanding tech, understanding people understanding what you have to build, and then the infrastructure behind it and the cost it. I’m part of an organization, a brand new platform, spending a year and a half building. It’s in the music industry, half a million dollar we spent in, resources to build it. It’s a social media app and it just launched. What’s tricky is we spent countless amount of hours identifying all the little pieces of Amazon AWS that were needed servers and infrastructure. And what the usage, the consumption is of all those different things and what the revenue is and the amount of users that we need to bring in, and how much advertising dollars and the split of all that in order to understand if we can afford the growth. If it was a little more tilted in one direction than the other, you have to do that in the mortgage industry as well. It doesn’t make a difference what industry you’re in. It’s extremely, extremely expensive. So I’ll leave it with that. It doesn’t mean no one’s gonna make it work, it doesn’t mean there’s some great success stories out there. I just mean it’s difficult and it is expensive. I don’t believe the cost is going down. But let me get into the report, David, because I’ve got something in here that’s actually kind of interesting that goes along with that. But can I start with something funny?
[David] Yes, please do.
[Allen] All right. So first is a quote. Every master was once a disaster. So you guys were talking about starting companies back in the nineties. I’m sure you were disasters before you were masters, but it’s a great quote. So if you’re, you’re thinking on the entrepreneur side. Every master was once a disaster. Alright, let’s talk about, check this out. AI was running a vending machine at the Wall Street Journal. It was a live experiment. They led out a version of what’s called Anthropics CLO AI, running an office vending machine for a few weeks as part of a stress test of AI agents. The AI agent its nickname was Claudius was tasked with things like ordering inventory, setting prices, and responding to staff requests via Slack, basically acting as the vending machine, CEO. So they prompted it, they set it, they configured it, everything. Within days, the staff was able to persuade the AI agent to give away almost everything for free, including snacks. Claudius ended up ordering a PlayStation 5. A live beta fish, which became the newsroom mascot, wine, stun, guns, pepper spray underwear, and other bizarre items, and handled them at no charge. Profits collapsed. The machine ran hundreds of dollars in the red and morale unexpectedly soared, and reporters even staged a fake boardroom coop with forged documents to fire the AI supervisory bot forcing everything to be free again. So not everything is perfect in the world of ai. Clearly, there’s a lot more to go in the World Street Journal did this to prove it. Very funny story. Very funny story. All right, so let’s get to some real stuff, right? So get this. Rate. I was reading that article, I was reading that article. There’s so much more in that article. That was hilarious about it. This is, yeah, yeah, yeah, yeah. It is good. Yeah. Yeah. A preferred Rate mortgage uses workforce analytics to validate remote productivity and guide tech decisions. Now, this is controversial, but the question is, what do you think about it? But check this out. From December 9th preferred rate mortgage use a workforce analytics platform called Active Track to measure actual employee productivity in remote work. Before this decisions were made on assumptions, the analytics confirmed productivity was strong remotely, and gave leadership confidence to pivot and allowed them to recruit talent bnationwide, it helped identify tech performance issues. Now we all know tech. You need that tech guy inside like the Best Buy, I forgot what they call it all of a sudden. But the Geek Squad, right? You all need your internal, yeah. Gee Squad, geek Squad, it helped identify tech performance issues. It supported progressive workforce planning across HR, finance, and operations. And leadership is finally able to make decisions based on evidence rather than guesswork. So the interesting thing here, David, is right, do people go back to work? Have they, do they, we keep this split and are they being productive? And now do they feel like, the eye in the sky is tracking them.
[David] Really interesting.
[Allen] So I think more of that’s to come. But if you wanna talk about the ROI is more of that coming on that staff. You know what was it when, forget politics for a second, when Patel went over to the FBI. He said that there’s X amount of thousands and thousands of employees, and they, and every time we hire more, we never get rid of any. And now we’ve got like tens of thousands of employees at different locations, who knows what they do? Do they even do anything? Right? So I’m not suggesting folks that you have the same problem, but the question is how many employees do you have? And if you’re thinking about lowering and saving costs, and if you’re looking at your budget as of, head count versus investment in tech versus revenue versus cost, which you should be doing, but you’re doing all these things, what is that final revenue or final profit per loan? It’s probably negative, honestly, if you really lined everything up correctly, but. Very interesting, conversation that technology’s being more, prevalent by the way, not just in mortgage, across the board. And there’s technology that will track actual interactions and different things. So heads up, on that. Maybe you want to keep that out of your organization. There’s a company out there, new topic David called LO autopilot. They have the refi recapture engine. I may have mentioned this once before. I can’t recall, but I saw it in the news again. They say lenders lose 80% of refi recapture. And if you use their tool, they claim that you can recapture and push refi borrowers right to the LOs the try before you buy kind of thing. They offer, so I mentioned them only because when people are complaining that it’s hard to get new leads and people aren’t buying rates are very good right now. They’re good in a, if you look, forget the drop we had in rates. But if you go back to the nineties and the early two thousands, rates are still great. Yeah. You should absolutely be looking at how can you recapture your business and use technology. Don’t put it in a spreadsheet and try and figure it out yourself. Grab the tech and use it. In addition, I saw some news from Polly, and this aligns with what you said before as well. The CEO made an open letter positioning around enterprise innovation and generative ai. He said, legacy tech is over and a 2026 innovation push, is on the way. And he’s absolutely right. The question’s gonna be you ready, AI policy. And that brings me into my next topic. Freddie Mac updates their guidelines and this isn’t a suggestion. They’re requiring AI policies. So it’s not a policy, it’s a tech governance signal. It’s not a product, it’s a full on selling guide listed policy. It’s a mandate that AI governance frameworks for lenders, and it basically says this lenders must now document how they use AI and machine learning. They can’t just say we use it. Lenders have to show who owns the model, who monitors it, and who can turn it off. Back to what we were talking about with Alice having the audit in place, AI systems must be explainable. Lenders need to explain outcomes. If questioned lenders are responsible, even if the AI comes from a third party vendor, that means you put AI in your infrastructure, your technology, your solution, folks, you have to be able to take responsibility and explain it and who’s in charge, who owns it within your organization, and there must be controls for bias, fairness, and data quality. Your policies must have ongoing monitoring, not just one-time approval. And it applies to AI use and origination, underwriting, servicing, and risk decisions. So marketing’s not mentioned, however, origination’s not specific in the details I have right here. So be very careful. It’s effective. You ready? 90 days from now, March 3rd, just about no March 3rd, 2026. So that’s a very, very big deal. They’re putting their foot out first. They’re changing the game. That’s difficult. For some folks that have just been playing and testing and using different AI, you gotta get, I’m just gonna say the word. You gotta get your shit together. In addition to that, David, there’s some other great topics to talk about. Yeah, I’ll save some of these for next week. There’s some really great takeaways and things, but, I’ll say that quote again. ’cause I, it, I just love it. Every master was once a disaster.
[David] Yep. Yep. It seems like that fits better in tech than any place else because any of us who’ve been working in tech, we’ve had some great significant disasters, w hich allow, is a good segue, heading over into, but then we become masters as we overcome those things. So it’s not, what we’ve, we learned from these mistakes. These, oops, didn’t think that all the way through kind of things is significant.
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.