[David] Let’s get over to Allen Pollack with a Tech Update. My friend. I’ve got a new set of headphones on here right now, Allen. They’re AI, they’re called Open AI headphones. It listens to everything, makes notes of everything, and I am so amazed. So if anyone looking at this, you can see these things are a little small. They sit right behind your ear and it is AI it’s making notes of anything and everything. Kinda help work organized My day. I might suggest. That’s crazy. I looking, I not 75, I don’t think I am, possible I may have a few brain cells falling off, but Alle.
[Allen] Sounds amazing. There’s all kinds of cool tools like that. I know that there’s a lot of kids, I’ve seen ’em on TikTok. , It looks like a battery backup device. Those small ones you can attach to your phone and they just listen to your classes and record notes for you and, make follow ups and all that stuff. So people are doing those things. I’d expect that we’re gonna see a lot more of that, but let’s talk about some AI stuff and also a comment that Bill said last week. So let’s start off with first NBC. It was the opening week of football. They tried an AI generated cartoon for the Eagles versus Cowboys game in their broadcast openers, and it was complete with player car graphics and mascot characters and fans were not happy. They took the social media. Social media was lit up. This is what the news article said. It was lit up calling the visuals embarrassing and disgusting over accentuating body pigmentation looking strange. Some people said, with the money you have, can’t you hire real animators? So at the end of the day, NBC I don’t know if they’ve responded to this, but the football fans that saw this went to the internet and were not very happy about it at all. And then second, this is interesting. In Ohio, a TikTok user endured a creepy job interview with an AI recruiter. Lemme break that down. Somebody got on the phone for an interview and was talking to an AI recruiter, so it was an automated recruiter. That got stuck repeating vertical bar pilates 14 times in 25 seconds. So real life distortion or broken code. Surely one of the most unsetting chatbot auditions ever. AI is everywhere. You never know when you’re talking to AI. Some of us can tell, some of us can’t. But if vertical bar Pilates gets mentioned 14 times in 25 seconds you know that you’re not talking to a real person. Alright let’s talk about Bill and what you had said last week. You said build versus buy and specifically you said just because you can doesn’t mean you should and there was an interesting quote that I found that I wanted to mention and we’re gonna get through a little bit of news and then get right back to the Bill’s buy comment. So the quote is, owning your code isn’t the same as owning your customers, and that’s really critical as you think about that build versus buy comment. Alright, let’s talk about a couple things going on in the market and then we’ll come right back down and talk about build versus buy homes. So it’s a company called HOME’s. They have what’s called a wholesale search. It’s a bilingual AI tool for lender matching. So this helps the Hispanic market. It’s called the Hispanic Organization of Mortgage Experts. And it’s a chat bot power. It’s actually off GPT powered bilingual AI tool that instantly surfaces lender requirements from 150 plus lenders and which is critical for borrowers lacking social security numbers, credit scores, or traditional documentation. And it speeds up the decision making and bridges that big information to access gap and boosts services for the underserved Hispanic communities. It went live this week, or actually four days ago, so within the last seven days. So check it out. That’s HOME. Blue Sage just launched an AI sales agent. It’s an AI powered digital lending assistant with their digital mortgage platform and handles borrower scenarios, loan lookups, and email drafting in real time.We’re gonna continue to see more of that. So hats off to Blue Sage and get this, Tech Builders show. They’re showing, or they’re seeing, and I’m, I don’t know this conference, but there, there is a tech builder show. I don’t know if it’s a show or a conference, maybe some of you out there do, but they’re showing a rise in DIY momentum in the mortgage industry. They’re seeing a growing trend revealed by National Mortgage Professional news shows lenders like Better Absolute Home Mortgage and Sunwest are building their own internal POS, LOS and AI tools and it matters because the move now reflects frustration with legacy platforms. Okay? So you get agility control and customization and going right into what Bill was saying. Let’s talk about this really quick. Some of the big folks that are out there, they have to own their own destiny, right? These are banks or financial institutions that have way over a billion dollars in assets. They have large infrastructure, they have old systems that they need to integrate to. They have massive reporting requirements. It’s not so simple for them to leverage everything that everyone else has. So you can’t take the example of what you’re doing and apply it to yourself, even if you are a billion dollar mortgage origination institution. It still doesn’t mean that. Let me break it down for you a little bit here. First, Open AI’s, SDK integrations. We all hear about it. We all think about it. Everybody talks that they have them. Open AIs mean flexibility. Okay? You can plug right into third party services without locking into that vendor APIs and SDKs. The SDKs is getting attached to the APIs. It speeds up innovation. You don’t reinvent the wheel, meaning you’re not having to code around all the APIs and you have shared code with the vendor. They’re providing you these SDKs. You build them into your system together those two, there are faster deployment and you can get new features across LOS and POS systems and CRMs, and then you have lights out automation. Okay, lights out is the vision. True end-to-end digital mortgage. Now, why are people talking about build versus buy? When you’re buying, you’ve got a proven vendor platform. Faster time to market. It’s guaranteed or hoping that it’s within the guardrails of regulatory and compliance. But you have less control. If you build, you have full control, you have differentiation, which is important to many. You have high cost of talent in long times to build that tech, and you have risk of being obsolete of what you built based on the market changes. There’s a hybrid approach a lot of people do. The trend is a lot of people want more of that core tech. They want some ownership, they wanna manage the reports. They’re tired of putting in change requests and paying for everything. But the warning that I give you is the build efforts are way underestimated. And just because AI’s out today, it doesn’t mean anything different. I’m gonna give you an example, a really quick example of the. Let’s pretend we were to build A CRM, you were to, and what the cost impact is. But let tell you what GPT said. Two quick little jokes about this build versus buy. The first thing it said is build versus buy is the new hot debate. I saw one lender proudly say we’re building our own tech stack, which basically means they hired someone’s nephew to wire up Zapier and the set, which is true. The second joke that GPT said is, everyone’s chasing that lights out automation, but in most shops, lights out just means the IT guy went home at five o’clock and nobody else can log in. So now let’s, which it’s funny, David, right? But it is the truth. How many of us have been worked at a lender where there’s a tech problem and there’s no one to talk to, right? Yeah. Lenders did not sign up to become technology companies. You want to pick up the phone and call, make someone accountable. Now, some lenders have implemented a staff that’s in charge of managing the vendors and they’re trained well on those vendor technology platforms. But it doesn’t mean you’re a technology company. Here’s how I’m gonna end our little segment today about technology, example of CRM, cost impact build versus buy. If you were to go build your own CRM, right? There’s a lot of tools out there in the world. You’ve got Zapier, you can get your own database. Two senior engineers. Okay? You can’t have just one because you need a tester and you need one sick. You need to have someone to back them up. You have two senior engineers, part-time support, a PM meaning that you’ve got someone to manage the project as well. You’re talking about 350 to 500KA year ongoing to have that staff. You’re still 12 months before you get anything even usable into the market if you were to start today, and there are hitting costs of compliance updates, your LOS and POS integrations, user adaption, training bugs, you name it. Okay? Half a million dollars a year. Let’s say if you were to just go to any one of the CRMs in our industry, on average with a SaaS pricing model, you’re gonna pay 50 to a hundred dollars per user. Meaning the people that you have that use the system, 200 users, let’s say, is gonna take you, it’s gonna cost you about 120 to 250 K a year. But you’re live in weeks. The integrations are known. They’re matured and tested. They may not be perfect, but they work, they’re tested. You can click right in and use them, and there’s a support staff built in and there’s innovation that always comes with it, and you can pay for customizations and changes. I know the vendors are probably all standing up clapping for me right now. Just like Josh Allen won the game last night, Monday night. Football was an amazing game, by the way. But the point I’m trying to make here is it’s gonna cost you a ton of money, anguish, stress, everything you can imagine with everything else you’re doing at your company as a lender to try and start building your own technology. Your team’s gonna get bigger, it’s gonna get more confusing. You’re gonna have staff issues, you’re going to have turnover. You’re gonna have issues with liability. If things not working, data breaches, the list goes on. So that’s my quick opinion. It doesn’t mean that’s the right thing for everyone, I hope everyone got those details and open for discussion. David?
[David] I think the world of technology, one of the thing that we’re doing now, listeners we’re doing an AI update every week. I got Pavan as a guest on there and we were talking about spaghetti code versus, real code. There are so many companies out there, you can now build your own code, build your own software. But you hit on the that point and we talked about that and go into that in greater depth. The continuity of code when it is written by humans versus the spaghetti code that can be written versus using these agents that are out there it’s just something that’s fraught with problems.
[Allen] Look, you bring up a good point, David. So I have a technology platform that is live in the market and it was built on an AI system. We tried it out and we liked it and we launched it. On Friday. It stopped communicating with an API that it needed to work with. We spent, I’d say a good 10 hours talking through, it’s an agentic experience, meaning you’re chatting on the left and it’s making all the changes on the right and you can watch it change all the code. It was using the wrong end points and it took a human to actually figure it out. The human went to ChatGPT separately and said, here’s the problem we’re having chat. GPT said, are you sure you’re using the right endpoint? But that was after hours and then went back to the AI software builder and said, Hey, did you look at this? And it came back and said, oh no, I didn’t. Let me try that. And then it worked. So it’s very easy for an AI platform to not to have too much information and make the wrong decision, and you still need a human in the mix. So what I’m getting at is the bigger and more complex of a system you build, even with AI, the more problems you’re gonna have. And you have live customers, right? You have your reputation on the line. This isn’t some little project or hobby you’re working on. So keep that in mind.
[David] Good point.
[Alice] I just wanted to jump in on a couple things that Allen in your assessment that you were comparing. I think one of the most important things is the user count and when you’re in a company and the avenue that you’re looking at. Loan origination is very different from one company to the next. An LOS that’s customized to fit how you do business, I think is an interesting use case as long as you’ve got a big enough user pool, right? You’re a company that’s 1500, 2000 people not sitting in a few hundred. And therefore you start to get that a better scalability on what that cost is and how agile you need to be, which you use that word. So I think that’s an important thing and which channel you’re in. So servicing, for example, servicing goes a certain way, right? You follow a certain set of rules. There’s great servicing systems out there. It’s that LOS that I think is something that really gets people’s interest or the POS to try the point of sale systems or the loan origination systems that people are really, I think taking a look at, should I build my own? Because that’s a shorter list of vendors out there and you can get really tied into a few of them and that you wanna move at a different place or a different direction than they do. And that is that’s the flexibility and control I think a lot of lenders are starting to look for. As you mentioned this is becoming a topic for those in that larger group.
[David] Good point.
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.