The Hitchhiker’s Guide to AI in Mortgage Lending with LenderLogix

AI has officially entered the mortgage universe, and most of us are still trying to figure out where it’s actually taking us. There’s a lot of noise, a lot of bold claims, and not a lot of clarity on what’s genuinely useful versus what just sounds impressive.

LenderLogix CEO Patrick O'Brien provides a practical field guide, breaking down what AI actually looks like inside mortgage tech today, where it’s already driving efficiency and improving borrower experience, and how to spot real value inside the systems you’re already using. 

He also talks about why personal experimentation matters, how teams can move off repetitive busywork, and what skills will matter as things keep evolving. No hype, no panic, just a clear way to think about where AI fits into your day-to-day and where it’s headed next.

Originally aired live on 4/14/2026

Transcript

Copy Transcript Copied

SPEAKER_01 0:05

Hello everyone. I'll give it a few seconds for people to join, but um we'll gouch with the OMBA and I'm joined by Patrick O'Brien, um CEO of LenderLogics. Um and he's gonna kick things off talking about mortgage lending and AI. So without further ado, I will leave it off to him.

SPEAKER_00 0:24

All right, appreciate it, Will. Hitchhiker's Guide to AI and mortgage lending. Um so we wanted to put together a presentation because we've all been thrown into this AI world and nobody really fully understands it yet, right? So kind of the hitchhiker's guide is this sort of cracked uh version of uh talking through what's going on and and how we're able to navigate this uh this new AI world that we're in in the mortgage space. Um and like Will said, my name is Patrick O'Brien. I'm with Lender Logics. And why we think we're well positioned for this conversation is because we're actually building and shipping software, AI solutions into the space. We've we've been building software for 10 years, and um, you know, we've um as as AI has leaned into the space, like we're actually shipping solutions. So we feel like we're in a good position to um, you know, kind of talk about where things are at, what we're hearing and and experiencing uh with our customers. So first thing, don't panic, right? Like nobody knows where AI is going in this space. You have many talking heads, you have many consultants and analysts and different vendors, and nobody really knows where AI is going. We know it's got a lot of um potential and there's a lot of energy behind it, but there's a lot of um, you know, claims and confusions, and you know, just a lot of people with with agendas that are kind of pulling in every direct direction. So what we wanted to do here was just kind of talk about the state of things, not this is not a sales pitch, this is not um, you know, this is not hype, just a framework for how people in the mortgage business, maybe as we start to think about how it's gonna impact our business, how it's gonna impact our careers, and just kind of an honest conversation about um where things are at, so we can kind of um you know react accordingly. Um everybody knows that the answer to everything has been AI, you know, so AI can do a lot of things, right? We can um efficiently process documents and add conditions to loans and um completely reimagine how we um you know reach out to our customers. But knowing the answer isn't enough. What's the real question? And and those are you know the the questions that we ask ourselves and the the the problems that we're seeking to solve in our businesses um is is the more important um component of that of that equation. So yeah, we get it, AI is the answer to everything, but uh you know what's the question is the uh is the more important uh piece. So uh chapter one here, is AI overhyped or is it transformational? And the answer, of course, is both. Um skipping ahead here. So the in order for AI to be impactful in our businesses, um, it needs to solve real problems. So if AI for the sake of AI does not matter. Um AI as a way of solving real business problems is everything. So, what are the types of bit problems in our businesses that we're seeking to solve? Number one, we we need to save time, we need to process loans uh more efficiently. Saving time for an individual means saving labor and driving the cost down. Workflow efficiency. Do we have um you know a million different tools not working together? Um, and uh process for the sake of process and um getting in the way of us getting loans you know closed quickly. Um, what's the revenue impact? The cost to originate a loan has uh ballooned over the last several years, and um everybody's in this constant quest to figure out ways to um drive that cost down. How can we take maybe more um labor costs, more vendor costs out of um producing a loan? So um, you know, increasing uh the throughput of loans, so how many loans we can get in the door, and then driving down um the cost per loan. And you know, headcount is is a component of that. How many, how much are we um paying folks to do kind of clicking and dragging around versus um you know, real impactful human work? So um looking through the lens of the problems that we're trying to solve in our business and potentially looking at AI as a way to solve those business problems is the right way of framing it versus AI for the sake of AI. I um, you know, you see some uh you see a lot of AI solutions out there, regardless of the mortgage space. One of them is um you see a lot with voice AI. You know, do I want to get a phone call from you know an AI bot that's having a conversation with me? Um, I guess it's cool, like the ability that you can do that and the the fact that these um you know language AI models are so good is pretty cool. But did that solve a problem for me? Am I um you know, am I am I increasing my the likelihood that I'm gonna win a deal by contacting a customer uh with uh you know with a with a voice AI versus a human being calling them? So um and being able to answer those questions and identify particular uh business problems within our organizations is a key component of all this, and then determining if AI is is a solution to that. Um, you know, oftentimes will be, but oftentimes, you know, we just have to make sure we're asking the right question. So I think an important topic that's on top of everybody's minds, whether it's in the mortgage industry or not, is kind of the this career reality check in how AI could potentially impact our jobs. Is is is AI threatening um to take my job? And I think the the primary takeaway is that AI fluency is not optional. Being fluent in AI moving forward is not optional. Um Accenture uh in one of their recently recent quarterly releases announced they were laying off 10,000 employees. And the 10,000 employees that were impacted by the layoffs, it was determined that those individuals could not be effectively reskilled for an AI-driven workforce. So, you know, as part of the re-evaluation of their workforce, they determined that certain individuals were not capable of thriving um in an AI-driven workforce. So um, you know, making sure that uh AI fluency is um a core competent, core competency moving forward is certainly a component of uh making sure that we're you know kind of uh harnessing and uh cultivating the the right skill set moving forward into an AI-driven um workforce. So what does AI fluency mean? You know, a lot of times we think of uh AI and it's it's obviously technical in nature, but you know, AI fluency is is less technical than you might think. It is not, you know, writing code, it is not tuning models, it does not require having a technical background, it does not uh require understanding machine learning and neural networks and um you know the skills that it would take to be a data scientist or an engineer. AI fluency means knowing what is what AI is capable of doing, understanding what it can do and what it cannot do, understanding when to bring AI into your workflow, being able to look at uh the work that you do and and assess where AI can provide, you know, maybe some some where you can grease the skids a little bit. Um using AI to improve your output. So if we, you know, if if if we're responsible for generating, you know, if if I'm a processor responsible generate for maintaining a pipeline of 30 loans, how can I use AI maybe to increase that to 40 loans? Um and I, you know, looking at one's workflow and identifying the opportunities where you can apply um AI and being confident with it, being confident, like you know, you know, we were we were kind of talking before we got on the call about AI and when you can trust it and when you can't trust it. And um, being fluent in AI means understanding, you know, where you can trust it and where you cannot. So being able to adopt it and being confident um in its application. So um, you know, as the workforce kind of evolves, certainly there will be people who are leaning on AI as a way to um multiply their output. Uh, but certainly there will be people that you know don't uh hone those AI fluency skills and aren't you know using those tools as a way to increase their output. And you know, unfortunately, as uh as the world evolves and uh companies start to make decisions about um you know how how the workforce uh is evolving, you know, which camp do you want to find yourself in? So AI fluency is no longer is no longer optional. Um it is a skill that each of us uh need to work on. So um, you know, it is it it is not learning AI is not hard. It's just new. And that's a you know, that's that's something with a with with a lot of things. People think, oh, it's hard, it's hard. It's just new. It's just new. It's not um, you know, we're we're working with systems that we're not used to working with in ways that we haven't interacted with um with software before. Um, it's terms maybe that we're not familiar with. Um so it's important to keep in mind that, you know, as as we kind of approach our um attitude and our the framing of AI in our work, that understanding these things is not hard. First of all, nobody knows it's it's new, so it's you can't, you know, you can't have come from a long uh deep background of of AI. Uh it's not hard, it's just new. So um it does not require deep technical skills, but just the willingness to learn and understand these new concepts and then how to um you know integrate them in your workflow and look at your work workflow through the lens of um these tools being being available. Um so our our next chapter on our journey here is um how do we start becoming AI fluent? So um there was a few resources that I wanted to kind of point to and encourage um people to check out because it's not, you know, there's so many free and easy resources um that you can access to grow your skill set and become AI fluent. So there's a few that I included here, and of course, like you know, there's there's um there's a there's an arms race in AI now between anthropic, which is Claude, and um OpenAI, which is Chat GPT, and um Google Gemini. Um and it really kind of doesn't matter um for the most part which path you choose, but each um at least of these three um organizations, Anthropic, OpenAI, and Google, all have a plethora of free um resources where you can, for example, Anthropic Academy has a ton of free classes that you can take where you receive a certification. So if you're willing to spend the couple hours to go through uh the online um modules, you'll earn a certification. You can put that on LinkedIn, you can kind of point to it that, you know, are am I trying to be a software developer? No, but I'm understanding these concepts of uh AI, understanding how I can apply them in my workflow, and I'm signaling to you know the people around me that you know I'm I'm fluent in in AI and understand how to bring it in. So whether it's anthropic, whether it's open AI, whether it's um Google Gemini, um, you know, take a screenshot or take a picture of it, Google these um resources, spend a few hours on the weekend just kind of figuring out, like, you know, take the classes, understand it, know how to what what the terms are. Again, it does not require one to be highly, highly technical in order to bring AI uh into the into your workflow. Um all right, chapter four. In that same vein, personal AI use is non-negotiable. So just like AI fluency is not optional, how you become AI fluent is by using AI in your personal life. And it's hard, I would say it's much harder to learn the concepts of AI in a professional environment for a few reasons. Number one, you're oftentimes working with sensitive data, which um, you know, certainly you don't want to be uh putting that into just consumer grade AI, you know, chat uh windows. Um and you know, you it's it's just you know, we we maybe have less flexibility to experiment with things in our professional life uh than we do in our personal life. So um bringing these AI tools into our personal life as a way of experimenting and learning helps provide a tremendous amount of context that you're then able to bring back um into your uh professional environment. So personal usage is really, and I would say that for most of the folks that I know and the people that that I work with, um, you know, we've all experimented with it in our in our personal lives and and kind of just play around with it um as a way of building that foundation so you have context for how to um you know kind of how to bring things into your uh professional environment. So on the screen here, I have I have a few sort of like ideas, I guess, of things that you projects, if you will, of things that you'd be able to um tackle using, you know, free to extremely cheap uh pieces of you know, uh, you know, the free version of ChatGPT or Gemini or whatever. But just a few projects of things that that you could do to kind of like start to hone your your AI skills. Um number one, you could take a voice note. Take a voice note on your phone, talk to your phone and and just braindump some random thoughts for you know five minutes. Take that recording and put it into Chat GPT or Gemini or um Claude and ask it to distill those thoughts um, you know, kind of in into a summary. Um you start to see kind of how how I should talk to it, how I should interact with it, um, the capabilities that it has to kind of adjust that type of media and create a certain output. Um take your uh homeowners insurance or your your car insurance documents and um upload it to Chat GPT and ask it to help you write an email to other competitive insurance companies based on the coverage and uh premiums that you're currently paying in order to see if you can get a better rate. It's just an easy thing you can do to you know play around with it and kind of see what that output looks like. Um it's incredible what um AI is able to do with pictures and being able to watch videos. Um, you know, you could you could take a picture of uh I I uh I I I actually did this like I I had a my furnace uh wasn't working uh a couple months ago, and um I was trying to troubleshoot it and figure out what's going on. Of course, there's like lights blinking and all this other kind of stuff, and you know, it tries to start and it doesn't. So I actually took a video of it on Gemini and uploaded the video to Gemini. Gemini is capable of of actually watching a video and hearing the noises and things like that. And um I did that and it gave me the likely scenario, and this is probably what's wrong, and because this light's doing this, and it actually helped me troubleshoot it so I was able to know what's going on. My house burned down like a month later, but you know, other than that, it was fine, just obviously kidding. But um, but being able to, you know, troubleshoot things in your life by taking pictures and stuff like that, like you'll start to figure out what the uh capabilities within the AI tools are. Um, and an easy one is also like just taking a taking a picture of yourself on your iPhone and um you know prompting the AI to um create a headshot, you know, a professional headshot. And you'll start to see kind of, you know, maybe it's it's doesn't always get it right the first time, but you start to understand how to talk to it and the different um, you know, different uh ways of massaging the conversation in order to do that. So those are just some examples of things you can kind of do in your personal life on your own time, using tools that are certainly available to you, just to start to figure out, you know, how can I um, you know, kind of start to hone my um my AI fluency skills. So, you know, the whole conversation that I'm that I'm trying to frame here um always kind of brings us back to, well, what are we doing in our professional lives as mortgage professionals? And how does this new AI world kind of intersect there? So um I think it's important to remember, you know, AI is not a chat bot. The chatbot is is kind of the first iteration of AI that we're seeing um, you know, on the consumer side. Um, but it's not, it's so much more than just putting a chatbot on a website and having it um interact with, you know, whether we're interacting with it or a consumer's interacting with it. Um, quite honestly, it's you know, as mortgage professionals, I don't even think that that's even a good um use case for uh AI. I don't, you know, I don't know a lot of loan officers that want to hand their business over to a chat bot to interact with their borrowers. Um so thinking about where AI actually lives in our workflow and you know, what what we're going to see in in the near term, certainly within the next few years, is the AI that we are going to interact with as mortgage professionals will be delivered through the systems that we're already using. And you already see this today. Um, so whether it's the LOS, the point of sale, our CRM, um, how how those systems might be handling um documents and automating tasks, um, you're starting to see AI being introduced through those systems. So these aren't um new, you know, these aren't new companies that are being introduced to us. Some, you know, that some of that exists certainly, but how most of us are going to interact with AI in the next few years is through the systems we're already using, through our LOS, through our point of sale. But being AI fluent allows us to recognize and know how to better activate those components of the software that we're using to be more efficient. So I think it's important to be mindful that, you know, if if we are not using the systems that we already have effectively, AI is not coming to save us. AI is not going to come from the outside and just completely reimagine our workflow. What we um what we need to do is understand how to best activate AI within the systems that we have because all of us that are on the the software side of the house are bringing those capabilities into the marketplace. But for the people that aren't using the software they're trying to do things manually they're handling everything over email aren't able to take advantage of the of those um those efficiencies. So be very mindful that AI is already being delivered to you through the systems that you already use and being aware of that and really trying to become expert in the systems that you have and recognize those capabilities is very important. So again, you know your CRM like we've been using CRMs for 20, 30 years in the mortgage business. How a CRM is able to automate campaigns, pull in data, you know, score leads, things of that nature, that functionality is is you know stepping up significantly because of the AI that's woven in point of sale as well, being able to triage documents and automate tasks and things like that. Weaving AI into those systems is just expanding the capability. The problem is if I'm not using that, if I'm not seeking to become a better user of the software that I have available that my company has put in place of those efficiencies that the AI is driving in. So if if you look at the uh the systems that you use and you don't think you know you're a maybe a power user um you know part of being uh preparing for the uh evolving workforce um means doing a better job of uh becoming expert on those systems so um you know poor adoption is going to result in poor outcomes like AI is not coming to save you um but it will um enhance the people that are you know already um you know uh working in those systems and becoming um more proficient um again ai is not just flash it's there are plenty of cool things I love watching demos and seeing all the cool things AI can do but um most of many of that is irrelevant to like business outcomes and what we are trying to accomplish does it save time does it drive revenue does it reduce cost does it improve my borrower experience again AI for the sake of AI if I were to bring in something that's deteriorating my borrower experience just because it's AI did I really accomplish what I wanted to accomplish probably not so I want to um recognize areas where it can improve and certainly areas maybe where bringing in AI isn't going to improve one of those things and I'm just looking at AI because it's flashy and cool um AI should allow us to be more human it should allow us to focus on our highest best use so um you know not replacing our the the the the humanity that we bring to a transaction which in mortgage is really a really important component um it's more about thinking you know what can I do with AI that I wouldn't be able to do otherwise how can I kind of almost be like superhuman is important. So you know as you as you as we look at our workflows and think about like what are the things that AI should do um and and that it's probably better at doing than than a human's doing you know formatting you know typing and formatting documents um you know clicking and dragging around systems um you know chasing conditions and classifying documents um a lot of data entry like you know those are things that that in this business a lot of human hours are spent on but if we can reduce the amount of time that we're spending on those tasks and allowing the AI um functionality that's coming into the business again usually through the systems that we're already using if we can let that do more of that lifting where we're able to focus on relationships focus on communicating focus on advising creating trust with our borrowers um judgment I mean the the the judgment of a mortgage professional to a borrower that's um you know a lot you know unsure and kind of seeking that um you know advise you know that the the advisory of a of a person a professional is so important so if we can um leverage ai to do to take some of those other tasks off our table um and drive more of our time spent towards actually talking to people communicating with them being empathetic for their situation helping them uh you know uh you know resolve whatever it is they need to resolve in order to buy a home those are the things that we're good at that you know we're not likely to see in the near term AI tackling um there's this uh uh great um scene in the hitchhiker's guide where um the earth's about to be dis demolished and everybody's kind of like shocked by it and the guy says well the plans were on display nobody looked at them I don't know they've been out there for years um and I you know I think that's kind of where we're sitting right now in the um in the AI space the the the hyperspace expressway is being built now and resistance is useless um technology moves forward it always does it always has um and it it solves problems in our business and I think as um you know competent individuals in the space it's important for us to realize that that's happening not resist um because technology always moves forward regardless of whether or not we want it to or not um you know two groups of people that are that are emerging those who leverage AI and those who don't and the gap will continue to widen and in this mortgage space you know as we start to look at our workflows and try to figure out you know what camp we find ourselves in um you know honing those skills for AI fluency bringing it more into our personal life will kind of help carve out which of those two groups um we find ourselves in so um nobody knows right like we don't know to what extent uh AI will continue to um disrupt the mortgage space uh but we certainly know the people who develop AI fluency who are using the tools um who are investing in the systems that they have access to and determining how to activate AI into those systems uh those will be the the folks that will kind of be prepared for the next generation of um you know of this mortgage workforce the evolving mortgage workforce so just remember don't panic it's not hard it's just new it's not hard it's just new um learn the tools um get fluent in the tools and you'll be just fine so half hours up appreciate you guys joining us yeah Patrick do you have like a a few extra minutes stay on sure um so we had a uh question in the chat um on your thoughts on cybersecurity so I'll give you like a a second or two to gather thoughts I also wanted to just comment on Rob's comment um saying that AI has kind of been around um you know for 25 30 years.

SPEAKER_01 31:30

Totally true. I mean Hitchhiker's Guide to the galaxy came out in like the late 70s and they talk about concepts in that but and I'm sure you'll agree I think it's more about like the mass widespread adoption of it um we've seen in the last few years. Like even the Roomba like the Roombas I think they use some AI but like we're now at the point where no matter what industry you're in even if it's maybe not going to be adopted widespread people are trying to figure out how they can and so I always thought I read something and it was you know it's not going to technically replace everyone's jobs but someone that's using AI versus someone that's not in the same role that's like the difference. And I always thought that that was like a pretty pretty good point on that.

SPEAKER_00 32:08

But um 100% yeah I mean you think back to automated underwriting and you know we've we've we've had um you know out of people bragging now about how you know you know you can underwrite a loan in 30 seconds. It's like well you could always do that with automated underwriting for the last uh 20 years. So again a lot of those concepts aren't new it's just kind of like the mass uh availability that the cost has been driven down so much. So just we're starting to see it introduced a lot more into our workflow and uh paying attention to it and um you know being ready for it versus fighting it I think is kind of the uh the important undertone.

SPEAKER_01 32:45

Totally agreed and do you have any thoughts on cybersecurity?

SPEAKER_00 32:47

Totally okay if not but I mean you know the you've you we've seen a lot out there um you know I think uh a person just straight up interacting with a consumer grade AI um take it for what it's worth and you you certainly do not want to be you know we we hear stories you know somebody's like oh I put somebody's pay stub in ChatGPT or whatever to compute the income and it's like nah you don't want to be doing that um you know as um technologists you know we have access to and you know pay a lot of money for quite honestly um tools that allow us to kind of access air gapped models and um you know uh follow uh you know a certain technical protocol. So you know at least us folks that have been in the the industry for a long period of time, you know, operate that way just as uh as table stakes. So um you know whereas maybe there's new entrants into the space who don't understand the regulatory framework and um you know kind of the vendor management side of the house, you know, I can't speak for everybody, but certainly the again the AI that you're starting to see in our space a lot of times it's coming from existing players that have been doing this for a long period of time, you know, have certain capabilities that go far beyond what what the consumer would have available to them in terms of uh a secure um AI environment.

SPEAKER_01 34:14

Awesome. Um and just uh Teresa I will uh talk with Patrick um offline after this call and if um we have like a little deck um copy or just some key messages I will send that out to everyone on the registrant list. Um so just stay tuned for that and uh thank you guys all for your comments really appreciate it. Um I think that concludes the webinar thank you again for your time and see you next webinar is next Thursday. So stay tuned on that. Thank you guys.