- 10 hours ago
Mortgage lenders are facing a structural challenge: rising costs, declining productivity, and increasing operational complexity. In this conversation, we’ll explore why AI has shifted from a promising innovation to a true leadership imperative—and how forward-thinking lenders are using it to transform their operating models without overhauling their core systems.
From automation and governance to change management and long-term economics, this session will outline what it takes to move from experimentation to real, enterprise-scale impact.
Because the industry is at an inflection point and most lenders haven’t figured out what to do about it yet.
If you’re responsible for operations, technology, or strategy, this is about getting a practical blueprint for what to do next, not just why it matters.
What you’ll learn:
Why AI has shifted from an experimental technology to a strategic imperative for lenders
Where to start: why mortgage processing offers the highest-impact entry point for AI
How to deploy AI by layering intelligence onto existing systems, rather than replacing your core stack
How to balance automation, human judgment, and oversight across complex workflows
What effective AI governance looks like in a highly regulated mortgage environment
Why change management – not technology – is often the biggest barrier to success
What will distinguish the lenders who lead the next five years from those who fall behind
From automation and governance to change management and long-term economics, this session will outline what it takes to move from experimentation to real, enterprise-scale impact.
Because the industry is at an inflection point and most lenders haven’t figured out what to do about it yet.
If you’re responsible for operations, technology, or strategy, this is about getting a practical blueprint for what to do next, not just why it matters.
What you’ll learn:
Why AI has shifted from an experimental technology to a strategic imperative for lenders
Where to start: why mortgage processing offers the highest-impact entry point for AI
How to deploy AI by layering intelligence onto existing systems, rather than replacing your core stack
How to balance automation, human judgment, and oversight across complex workflows
What effective AI governance looks like in a highly regulated mortgage environment
Why change management – not technology – is often the biggest barrier to success
What will distinguish the lenders who lead the next five years from those who fall behind
Category
📚
LearningTranscript
00:00:02Welcome, everyone. We are about to get started on today's webinar. I'm Allison LaForgia,
00:00:10the managing editor of the content studio at HousingWire, and today's webinar,
00:00:15rewiring mortgage for the AI era. Why AI is now a leadership imperative is produced in partnership
00:00:23with JazzX. A few housekeeping items before we get started with today's content. At the top of
00:00:30your screen, you'll see an option for chat, ask a question, resources, and even emoji reactions at
00:00:36the bottom of your screen. This is an interactive webinar. We want to hear from you throughout
00:00:41today's session. If you have any questions for either of our panelists today, you can submit them
00:00:47at any point in time by using either the chat or the ask a question feature. We will be keeping
00:00:53an
00:00:53eye on both throughout the conversation. We will be answering those questions towards the end of
00:00:59today's session, but we encourage you to submit them throughout the conversation. A recording of
00:01:04today's session will be sent to registrants early next week, so there's no need to worry if you miss
00:01:09something or want to rewatch a part of today's webinar or even pass it on to someone in your
00:01:15office or someone else in the industry. Now, today, we are joined by two experts who are going to lead
00:01:22us through this conversation about AI implementation in mortgage. We have Siddhartha Agarval, the Chief
00:01:30Executive Officer at JazzX, and we are also joined by Eric Hart, the President and CEO of Pulte Financial
00:01:36Services. Eric, Siddhartha, thank you for joining me today.
00:01:41Eric Hart, thank you. Now, I am excited to get it started with today's conversation. AI is at the top
00:01:49of mind for everyone in the housing industry, and I am excited to hear both of your perspectives on it.
00:01:56We're going to start with you, Eric. You've said that the mortgage industry is facing structural
00:02:02productivity and a cost problem, not just a cyclical one. What do you think is fundamentally broken in the
00:02:10current operating model?
00:02:14Yes, so it's a great question. I think probably most of the folks listening to this
00:02:19are familiar with how expensive cost of production and mortgage has gotten. NBA just came out with their
00:02:25most recently quarter statistics, I think $11,000 to make a mortgage. And it's actually higher in some
00:02:34segments, depositories, for instance. There's a really funny stat that it costs about $15,000 for
00:02:41depositories to make a mortgage, which is roughly the same cost that it is in the United States to
00:02:48have a baby when you factor in for somebody who has employer paid health insurance. So I think one of
00:02:57those should cost more than the other. I'll let you guess as to which one should, but there really
00:03:04is a structural cost challenge. When you look at other areas of consumer finance, credit card,
00:03:11acquisition, auto lending, personal lending, all of those, particularly the manufacturing of the actual
00:03:18loan or credit product, costs have come down mostly due to automation. Home lending has gone the other
00:03:24direction. I think there's a phenomenon in economics called the Baumol effect, something that was also
00:03:32called a cost disease. It talks about how in lower productivity industries or segments where you don't
00:03:41have technology driving innovation and higher productivity, you actually see wages rise because
00:03:48in order to attract talent from industries where productivity is strong and wages are rising, you
00:03:56just have to keep spending more and more even though the productivity doesn't get any better. And the
00:04:01classic example of that that they sort of did, these two Princeton economists did to articulate that was
00:04:08think about the idea of a string quartet. It takes a string quartet 30 minutes to play a Beethoven
00:04:15quartet piece. You can't make it faster. There's no efficiencies you can gain out of it. So over time,
00:04:21you have to keep spending more and more to hire those musicians because you can't get any efficiencies
00:04:27out of it. And, you know, if I were to extend that analogy, you think about the quartet in the
00:04:32mortgage
00:04:33space. It's a loan officer, it's a processor, it's an originator, and it's a closer. It's four people
00:04:38holding the bow on those instruments. And, you know, as long as it's sort of a human mediated process,
00:04:44it's hard to drive significant cost and efficiencies, improvements. And so over time,
00:04:52I think we've seen this space just effectively grow with the cost of labor, even though you haven't
00:04:58gotten any efficiency. So it doesn't get better, it just keeps getting worse. Those costs get passed on to
00:05:05consumers. And I think it's one of the big drivers of why home affordability continues to be so
00:05:12challenged. Absolutely. Now, I want to talk a little bit about why AI is becoming a strategic
00:05:22imperative in mortgage and not just an innovation topic. Siddhartha, let's start with you.
00:05:28Yeah, no, it's a wonderful question. I love Eric's analogy of musicians to the loan processing
00:05:35ecosystem. I would not have thought about that. But I think, you know, the thing that has changed
00:05:40most recently with generative AI is that AI is no longer just about, oh, it can automate tasks,
00:05:47but it's actually about reasoning, being able to think about things, being able to act and execute
00:05:54and get things done. So we're now moving from this notion of, you know, automating to actually being able
00:06:01to think about outcomes. And I think that is what is really driving why AI should be an imperative
00:06:08for folks to think about. So when you think about, you know, large loan packages, underwriting
00:06:14guidelines, lender overlays, or evaluating conditions, all of these things can actually
00:06:20now be done by AI. And if you think about evaluating conditions, you need to understand Freddie, Fannie,
00:06:26policies and guidelines. AI can actually do that. You need to be able to understand these large loan
00:06:32packages and map those loan packages and the data inside those loan packages into the policies and
00:06:40guidelines. So you can say, oh, this is why this condition is passed or failed. Now, all of this
00:06:46was very manual in the past was error prone, and caused rework of things. And now with AI, you can
00:06:54actually have this done in a more automated fashion. And at the end of the day, why does it matter
00:07:00to the
00:07:00business? And I just use that as an example, is because the economics change, the productivity of the
00:07:06quartet that Eric talked about changes completely, like the loan processor and underwriter actually
00:07:13doing the same thing, multiple times when they're doing a condition evaluation. Now, that can exist
00:07:20end to end, the cost structure changes because of the, you know, what has to be done by different
00:07:26people and how fast that could be done. Flexibility as interest rates might go up or interest rates go
00:07:32down. Now it becomes very easy to be able to scale up and scale down without necessarily needing to
00:07:38think about the humans in the process. And then responsiveness, you know, when you think about the
00:07:44front end of the business, which is the borrower, the loan officer, they're getting a completely
00:07:49different responsiveness, which means the conversion rates could go up, and your top line could go up.
00:07:54So those are some of the economics that makes it very important for people to think about AI
00:07:59first. And those organizations or lenders that think early about redesigning how the mortgage work
00:08:06gets done will probably stay ahead of the competitors.
00:08:13Eric, I love your perspective.
00:08:16I'd add on to that, Siddhartha, extending the metaphor, right?
00:08:23The humans have been intrinsically placed in the center of the mortgage loop.
00:08:30Originally because you didn't have any other way to do it. But I think even the current regulatory
00:08:34environment we have is sort of enshrined that there's human beings doing each role. They have
00:08:40an NMLS license and there are specific expectations where they sit, how many licenses they hold. It's all
00:08:48been designed around four people sitting in a chair, each one of them holding a very specific instrument.
00:08:53And I think that's going to be one of the really interesting things to see how it develops.
00:08:59The work is there. When you change how much of the work is automated and how much of it has
00:09:04to be
00:09:04intermediate by a human, you change the roles. And then you start to transform. What is the quartet
00:09:13actually? What instruments are there? What music are you playing? It looks entirely different. I think the
00:09:19technology is probably going to sprint ahead of maybe some of the regulators and their understanding
00:09:25of it. And that will create some interesting conversations. You know, it's funny. One of the
00:09:31things in, when you look in Dodd-Frank, which I think everybody's been talking a lot of,
00:09:36given that Congressman Frank's unfortunate passing of late, it doesn't actually anywhere mention in
00:09:45many of the parts of Dodd-Frank that you have to have a loan officer. It just says, here are
00:09:50the things
00:09:50the loan officer has to do and what the requirements are. I don't think anybody ever envisioned a world where
00:09:56some of these roles, um, don't, don't exist. Um, uh, there might still be people doing the work,
00:10:03but the role itself could look entirely different. Um, so it's, it's an interesting, uh, we're at an
00:10:09interesting inflection point right now.
00:10:15All right. Well, to our audience, this is one of those times where we want to get you involved with
00:10:23today's conversation. So we have a question for you. Where do you expect AI to have the biggest
00:10:29near-term impact in mortgage? Is it in processing, underwriting, closing, the borrower-Orello experience,
00:10:39compliance, and audit? Let us know what, where you think AI could have the biggest near-term
00:10:49impact? And we're going to give everyone a moment to send in their answers.
00:10:56And I'm not going to ask either of you to give us your previews so that we don't skew any
00:11:01of the results.
00:11:06Okay. The results are coming in.
00:11:12We'll give everybody a few more seconds to send in their answers.
00:11:22All right. If you haven't sent your answer already, now is your time to do so. Eric,
00:11:28what would your answer be?
00:11:33Uh, remind me of the question again. I'm sorry. I was.
00:11:37Where do you expect AI to have the biggest near-term impact in mortgage? Processing, underwriting,
00:11:43closing, closing, borrower, or LO experience, or compliance and audit?
00:11:49You know, it's, it's going to be a different question for different, um, lenders based on
00:11:55the nature of their business model. I think, um, you know, certainly the core production space,
00:12:01um, processing, underwriting, closing, um, it's a, it's a perfect,
00:12:08it's a perfect use case for AI where it's a combination of some structured data, a generally
00:12:14rules-based, um, kind of, uh, decision hierarchy. And, um, you know, but with the need to connect
00:12:23multiple pieces of information together using, you know, uh, using intelligence. I think that's the,
00:12:28the perfect use case for it. Um, I do actually think compliance is going to emerge as a real
00:12:35opportunity. Um, you know, we spend so much time and effort on the front end, trying to make sure
00:12:40that you manufacture alone perfectly. You have stops in your system, all these things to ensure that
00:12:46you don't make mistakes while you're manufacturing it. Well, I think, you know, AI is going to reduce
00:12:51the cost of being able to go through and do fulsome reviews end to end on a hundred percent of
00:12:56your
00:12:56files. So a world where you're, you know, your, um, post-production QC becomes much cheaper and more
00:13:03automated that, that changes the cost structure you have to put, you know, uh, pre-production. So
00:13:09those are two areas I think are going to have tremendous opportunities right out of the gates.
00:13:15Long-term, I think the front end and the human interaction part can, can, will be very exciting,
00:13:20but I think that one's going to take a little longer to flesh out, um, as we see how people
00:13:27where people get comfortable when they deal with, with AI from a consumer standpoint,
00:13:32on the back end user, I, I think that, um, the impact is going to be fast and furious.
00:13:37Now to my audience, you can see the results at the bottom of your screen, and I'm going to ask
00:13:42that Arthur, if he is surprised by these results. So we see processing and underwriting take the two lead
00:13:48positions with about 30%, uh, responses there followed in a close second by the borrower and
00:13:56LO experience. And by close, I mean, 4% close. So very, very close results. And then the next most
00:14:04popular result is compliance and audit. Does that surprise you?
00:14:09No, actually, you know, I think what's really, what's really interesting is that we have to
00:14:15think about this era of transformation as an operating model change more than a technology project.
00:14:22And when you think about the operating model change, I think the questions we need to ask is
00:14:28what task of the question we should not be asking is what task can we automate? Rather, we should be
00:14:34asking ourselves, what operational bottlenecks are we trying to solve? And then you think about the
00:14:41audience questions and, you know, the, the, the place where you want to address the operational bottlenecks
00:14:47first, where are places where there's high manual effort. So for example, the loan processor and underwriter
00:14:53are doing a ton of work, evaluating the conditions and mapping the documents, et cetera. There's complex
00:14:59document interpretation, both on the policy side and on the loan package. And there are significant
00:15:04operational bottlenecks in terms of how things move from one phase to another phase, from one player in the
00:15:10quarter to another player. And I think that is why, um, fulfillment, uh, with processing and
00:15:16underwriting are coming across as very strong initial use cases, because the processors are spending so
00:15:22much time reviewing documents, validating information, interpreting the guidelines, the guidelines are
00:15:27changing almost every month, you know, Freddie, Fannie are releasing things almost every month, and then
00:15:33evaluating the conditions, you know, imagine how much knowledge needs to be in their heads, to be able to be
00:15:38evaluate those conditions, and then move the files forward. So I think that because generative AI cannot reason
00:15:45over these things, and it can orchestrate workflows and workflows that are very long,
00:15:49it's not like they have to be done now, the workflows can be over weeks, or can be over months,
00:15:54etc. I think that is the
00:15:55right place. Like at Jazz, uh, JazzX, what we've done is, we've built a reasoner, and that reasoner can reason
00:16:01over loan
00:16:02packages, evaluate the condition against the policies, identify the missing information, and to Eric's pointer and
00:16:09compliance, explain exactly why the decisions were made, and give that provenance. So I think, you know,
00:16:16if business leaders are looking at prioritizing based on in business impact, repeatability, governance,
00:16:23readiness, you know, that's what, you know, will be the highest value outcomes, and think more about
00:16:30workflows that span multiple systems and roles, because that's where AI can improve the end-to-end operating
00:16:35model, not in a siloed functional area. Eric, I see you nodding. Do you want to jump in?
00:16:48Well, I mean, I think Siddhartha hit on a number of the key points, this idea of asynchronous work that
00:16:54can
00:16:54be done in the background while you're, you know, this, so much of, so much of the, the heartache in
00:17:02mortgage production is, how do you, how do you effectively build workflow? Every, it's a non-linear
00:17:12process for most loans, right? Yeah, occasionally you get the loan that comes in, it's very simple,
00:17:16and you ask for this, and you get this information, it moves to this step, and, and you can just
00:17:19move
00:17:20stepwise through, but in most cases it's, it becomes very bespoke. And so the efficiencies around, you know,
00:17:27how and when do you pull information in? How and when do you make decisions? How do you route one
00:17:33task
00:17:33from, from point A to point B? You know, a lot of the leading lenders in the space, the rockets
00:17:40of the
00:17:40world have through, you know, inch at a time over decades, figured out how to program workflows to
00:17:46try to predictively score, should I do this, then do that, and, and built a lot of efficiencies
00:17:52one inch at a time. And I think what we're seeing with agentic AI is you take a human being's
00:17:58ability
00:17:58to use their own intelligence to organically decide where, how you prioritize, and now you,
00:18:05you create agents that can do that, and you can have as many as you need. It, it, it opens
00:18:11up a,
00:18:13just a whole different way of thinking about the sequencing of work. So even if you're not getting,
00:18:19you know, automation, which I think there's tons of opportunity in automation, we have automation
00:18:25today, we have with OCR, there's lots of places where you can use automation. It's how do you ensure
00:18:30that the automation being sequenced in a way where it actually creates efficiencies. And that's the part
00:18:36where I think humans and systems have struggled to maximize that, and where agentic AI, I think,
00:18:43has the ability to, um, to supercharge that. So, um, uh, I agree with, with Siddhartha's, uh, broader
00:18:50point around, um, just asynchronicity of work being a huge opportunity. Yeah. It rules, rules-based
00:18:58systems are very brittle and reasoning-based systems, which are now a possible, those are not, you know,
00:19:06that, uh, you know, structured and this is exactly what you're going to do when this happens and you do
00:19:10this
00:19:10next, right? Because to Eric's point, exceptions are always happening in that mortgage manufacturing
00:19:17process, and it's how do you deal with those exceptions and yet be able to deliver a deterministic
00:19:23outcome that can be justified. Now, there's a couple of questions in the audience that I think
00:19:32we've touched on lightly, but I want to get an explicit start point out there. So where are the most
00:19:40practical starting points for AI in the mortgage process and how should lenders prioritize those?
00:19:46I think we just answered that, Alison. That's the question we just answered, right?
00:19:51Right. But I wanted to make sure that we just started with a simple answer because the...
00:19:57Oh, I'll jump, I'll jump in here, Alison. I, this is the, this is the, the best response to a
00:20:02good
00:20:03question, right? It depends. Um, no, look, I, I think it really does depend on, um, your business model.
00:20:10So if I'm a retail lender, uh, maybe a small broker, um, you know, so much of my time and
00:20:17effort is around
00:20:19getting eyeballs, converting those into applications, pulling them through, keeping them in the process.
00:20:25You know, I, I think, um, there's probably, um, a lot of opportunity to start with point of sale,
00:20:34um, top of the funnel, um, marketing, uh, uh, efficiencies so that that time and the money you're
00:20:42spending to get your business is, um, is efficiently spent. I think for, for, uh, larger lenders, those with
00:20:49baby built-in distribution channels, particularly banks, credit unions, um, you know, as, as a builder
00:20:55affiliate where we don't really do retail lending, we only serve the clients who are purchasing homes
00:21:00from, um, our, our home builder parent. Um, I think the, I think you start with, with, with production
00:21:06efficiencies, um, cause that's where the biggest opportunity is. And, um, and, uh, the dollars are there.
00:21:13I think you can learn faster. Um, uh, it's going to be an easier place where
00:21:19um, if you're working with a counterparty, they are going to see, um, uh, uh, transferable lessons
00:21:26that can, um, they can help you take advantage of as well. Um, and, uh, and then it'll create
00:21:34cost reductions you can use to pay forward into your pricing model, into, um, your service levels
00:21:40for customers. So, you know, I, I think there probably is a fork in the road for some lenders.
00:21:45It really is going to be about how do I, in an environment like this, where
00:21:49business is hard to come by? How do I use AI to make the most of that? Um, and then
00:21:53I think those
00:21:54that have, um, you know, a business model that's more focused on how do we execute the business
00:21:59that we do get? Um, I think, uh, uh, you know, core processing and underwriting is, is probably
00:22:06the, going to be the fastest road to, to making some impact.
00:22:12All right. Well, now there's, of course, the balance between lenders thinking about benefiting
00:22:18from AI while still preserving their existing systems and technology investments. How should
00:22:23they be thinking about this? Yeah, if, you know, um, and I'll bring my technology lens and I'm sure
00:22:30that, uh, Eric will bring his business lens. You know, I don't think one should think about AI as
00:22:36ripping and replacing, uh, if you have AI, then you got to replace your core systems.
00:22:41I think what AI is doing is introducing a new layer, uh, a layer that we might want to call
00:22:47a system of intelligence that can sit on top of existing platforms like the LOS's,
00:22:53but that system of intelligence can handle reasoning, workflow orchestration, document
00:22:58understanding, policy interpretation, et cetera. And the, you know, I think back to my Oracle days,
00:23:04and this is somewhat similar to what happened in the ERP world. You know, we used to have something
00:23:08about EBS, EBS, E-business suite. Um, one of the biggest challenges was that business logic became
00:23:15deeply embedded inside EBS. So customers would customize EBS and put all the business logic inside
00:23:21EBS and effectively change the DNA of that E-business suite application. So then over time,
00:23:28every customization made upgrades slower and took two to three years to do upgrades. It became
00:23:34more expensive and harder to manage. I think there's a similar pattern that is emerging in
00:23:38mortgage today where too many business, too much of the business logic is getting indoctrinated into
00:23:44the LOS or maybe fragmented across the point solutions. So I think the opportunity from AI is
00:23:51to decouple the intelligence from those core transaction systems and have a system of intelligence
00:23:57that's residing on top of your system of record where the LOS, for example, the servicing platforms,
00:24:03et cetera, are storing the transactions, maintaining compliance, et cetera. And, um, so AI, you know,
00:24:12uh, sits on top and it helps preserve existing lender investments, uh, and moving that business
00:24:17logic to that layer lowers the risk, improves the adoption and allows organizations to transform
00:24:24incrementally instead of, you know, trying to do a big bang, uh, replacement approach. So,
00:24:29you know, preserves the existing investment yet is able to leverage AI.
00:24:36Yeah, that's a really good point, Siddhartha. You know, I mean, what is a,
00:24:40if you're running a mortgage company, it's, it's sort of like, um, at least in the technology space,
00:24:45the metaphor I would use, it's like being the athletic director of a D1, uh, uh, college sports
00:24:51program. It all comes down to sort of like, who's your football coach, right? Or who, like that,
00:24:57that's the center of everything. And what is your LOS? What are you going to choose for your LOS?
00:25:02Um, and you know, what is an LOS? I said, Arthur, to your point, it's,
00:25:07it's a data repository, it's a business rules engine, and it's an interface that your team
00:25:14members use to transact. And that, you know, the, the data repository and the rules engines,
00:25:19I mean, I, I don't think they have to work. Um, they have to work in, in line with the
00:25:26requirements
00:25:26of the products that you're making, but all of the time, the money, the effort, the intelligence
00:25:30is built around the, the transaction interface, right? And that's where you, you know, it takes
00:25:36you months to train somebody in how to use your system. And here's how we navigate from this screen
00:25:40to that screen, which buttons we push. And, and, um, I think, uh, what agentic solutions can do,
00:25:48add that orchestration layer where, um, uh, you can continually upgrade and, and in some cases,
00:25:55customize those, those, um, interfaces to the needs of individual team members or their, or their
00:26:01portfolio of work. Um, and, and you don't have to worry about maintaining and managing that,
00:26:07that technology platform. Um, the, the metaphor I would use for how I think about
00:26:13our legacy technology stack and what we can do with AI. Um, it's a bit far afield here, but
00:26:19if you think about, um, kind of the United States military right now and the decisions they're having
00:26:26to make about how they invest in their future technology, we're watching in these horrible
00:26:30conflicts in Ukraine and in Iran, like the pace of change around, uh, the battlefield is, is,
00:26:38is remarkable. Um, you know, we look at drone technology and missile technology, it's, um,
00:26:44every single month, there seems to be some new innovation coming out. I mean, um, these folks in
00:26:49Ukraine, they're, they're inventing whole new weapons systems in, you know, in garages and putting
00:26:55them out to field a month later. Um, and so you, you might think, okay, well, that means company,
00:27:02countries like United States with these big legacy military systems, planes, boats, et cetera,
00:27:07they're going to have to throw all that out and move to something new. And what we're seeing is,
00:27:11is actually the opposite. Um, you know, last month, uh, um, United States, uh, air force, uh, put
00:27:18in order for like 300 F-15 jets. Um, that's a jet that first flew in 1972 and they are
00:27:25going through
00:27:26and doing a modernization on their B-52 bombers, which first flew in the Truman administration.
00:27:31And they haven't built since like the early sixties. And why? Because when you equip them
00:27:37with new data uplink systems, with better radar communication protocols, these old legacy systems
00:27:44turn into, you know, bomb trucks that you can equip with whatever the newest weapon is. And
00:27:49as I said, it's a far more consequential, uh, use case for modular technology than mortgages. But I
00:27:56think the same lesson is, um, there's value in a lot of these legacy tech stacks that we've built
00:28:01there. Um, they work well, they have traceable data, our regulators understand them. We, we, we,
00:28:07we know how they work. Um, being able to, to bring, you know, agentic AI and to supercharge
00:28:14those platforms with new capabilities, I think, um, you know, it gives you the best of both worlds
00:28:20over time. Look, all of us are going to be replacing those legacy applications with more AI native
00:28:26software tools, but the ability to do that incrementally to do it with, you know, thoughtfulness
00:28:32and change management that that's, you know, that's, uh, one of the most exciting things about
00:28:37AI. It doesn't feel like we have to throw all of our old work away and do something new. It's
00:28:42rather we can take what we have and, and, and make it better.
00:28:45I love Eric's, uh, metaphors. I mean, it's like metaphor after metaphor after metaphor. I need
00:28:50that skill, Eric. You're going to have to teach us all how that comes into your mind.
00:28:55That's right. They're not mine. I've got it. I've got a live clod feed next to me.
00:29:00Okay. That's what I need. A live clod feed. Okay. Got it.
00:29:04There we go. So we touched on a couple of things here. So to our audience, you're going to see
00:29:09a
00:29:10question at the bottom of your screen. Now that we've talked about getting started and balancing
00:29:15legacy systems while implementing AI or tech stack, what is the biggest barrier to deploying AI in your
00:29:23organization? Is it change management? Like Eric just touched on or started to introduce,
00:29:28which we'll get into in a second, or is it integration with current systems or is it trust
00:29:33governance or is it unclear ROI or is it executive management? I see our first couple of responses
00:29:41are coming in. So let us know what is your barrier before we get into our next section.
00:29:54The answers are pretty evenly split here.
00:30:00All right. Okay. Last five seconds to submit your responses and I'm going to let you guys see the
00:30:09responses. Ooh. Some people were pondering these questions. All right.
00:30:16Suddenly you get a bump up in the last five seconds.
00:30:19Yeah. Okay. Deadlines are always great.
00:30:22Deadlines are great. All right, here we go. So we see our top result, our top response coming
00:30:32integration with current systems, followed closely by trust and governance, and then followed by unclear ROI,
00:30:42then change management, which is tied with executive alignment, which gives us some good places to move
00:30:50on to. So Eric just touched on at the end of his question, change manage, at the end of his
00:30:57answer,
00:30:58change management. So I want to bring back that topic and start with you, Eric, to get your perspective
00:31:07on what lenders should be thinking about on the change management side of AI transformation. How does it
00:31:13actually impact roles, workflows, and the organization?
00:31:17Yeah. You know, I think there's two areas of change management that, you know, we've observed.
00:31:26The first one is, and you can almost say is this part of trust in governments, but it's how do
00:31:33you bring
00:31:34these solutions into your organization? How do you move towards, you know, towards even deploying,
00:31:41you know, new AI tools? You know, most of the innovation on AI we're seeing right now, unfortunately,
00:31:49is not coming from established service and product providers. They're doing some incremental things,
00:31:53but the more disruptive ideas we're seeing are from new players in the space who don't have track records,
00:32:00who are, you know, younger companies. They're more nimble, but, you know, if you're a large lender
00:32:06or bank, you know, you see, you see counterparty risk everywhere, and you certainly see it with
00:32:11newer, you know, startups. Add that to the fact that this isn't, these are new kinds of contractual
00:32:19agreements that you have to work through. Getting your arms wrapped around, you know,
00:32:22how do you share risk, data ownership? You know, as the, as the system learns, how much of that is
00:32:29your
00:32:29IP versus their IP? How do you cost sharing computation resources? So it's, there's a,
00:32:35there's a whole onboarding change management piece that I think we've seen is a, is a place that we're
00:32:41going to have to get better. And that's before you can get into now that you have these tools,
00:32:46how do you actually change within the organization to make the most of them? I think what we're seeing
00:32:52is that the biggest change management hurdles isn't a strong word, but the headwinds that we'll
00:32:58have to work through is the tendency to want to take these new tools and just apply them to your
00:33:05existing processes and your existing, you know, way of doing things. Great. This is just where we,
00:33:11I'll go back to the string quartet example. Great. It's still four people sitting in a chair
00:33:17playing this kind of music. And so now we've given them better instruments and just that's where we'll
00:33:23get the efficiencies. No, I think you actually have to deconstruct some of the roles and responsibilities
00:33:29you have, think about your processes differently. And, and that's harder for, for I think mortgage
00:33:37companies, which really do have these very clearly defined regimented industry standardized roles.
00:33:44That's going to be the hardest thing is to break those apart in a way that you might not see
00:33:49in a,
00:33:50you know, in other areas where, um, uh, you know, where it's, it's just not so cut and dry.
00:33:57This person does this and this person does that. So those, I think if you can, you can get through
00:34:02the change management, how do you actually partner up and work with some of these new technology and
00:34:07solution providers get past that. Then you got the, the exciting, but challenging work of how do you
00:34:14envision your workforce differently? Um, how do you think about the roles you have? And then,
00:34:20oh, by the way, how do you think about your agentic workforce? Um, uh, cause you've got a virtual
00:34:25workforce now, not just the folks that you're, um, you know, you're, you're writing paychecks to.
00:34:32Uh, you know, that last statement around how do you manage your agentic workforce,
00:34:37the digital workers that you have, I think that's wonderful. And, you know, I take a little
00:34:41contrarian lens to what the audience might have put out as the biggest barrier. Um, I really think
00:34:47that change management is very important in the AI world, uh, where the change management stream should
00:34:56run in parallel to the technology project. Um, because AI is changing how workflows happen. AI is
00:35:06changing how decisions get made. AI is changing productivity expectations. You know, for example,
00:35:13if you take an underwriter today, they process about 1.8 loans per day, and maybe different lenders
00:35:19might have different, but let's say it's 1.8 to 2.2 loans per day. Well, what if, uh, AI
00:35:26handles large
00:35:27portions of the guideline interpretation, evidence gathering, and condition evaluation. Suddenly,
00:35:33the human is operating at a very different productivity level. So what is the human supposed
00:35:38to do or what do you want the human to do? Then other things that AI changes is organizational
00:35:43structures and also how, um, institutional managed knowledge gets managed. And I think that that is
00:35:50probably the most profound thing is today a tremendous amount of mortgage expertise lives inside
00:35:56individual processors and underwriters. If you take an AI enabled organization,
00:36:01every time an underwriter overrides what the system or the AI system recommended,
00:36:07that now becomes a learning opportunity. So with that learning opportunity now, how do you decide
00:36:12whether this, what these five underwriters said, should that become part of the policy of the
00:36:18organization? This might mean you might need new roles. For example, a policy supervisor role,
00:36:25and does not have to be new headcount, but this is a subject matter expert to whom, uh,
00:36:30jazz is bringing all the changes that underwriters said yes, no to, and why they said that. And now it
00:36:36is this policy supervisor that's answering questions like, should these overrides become part of my
00:36:41organization's policy? Should we add this information into our knowledge, uh, into our knowledge base?
00:36:46Or should the AI be given these policy, uh, updates that can reason differently? Should we add this into
00:36:53the reason or thought process? So I think this, this, this whole notion of, you know, um, uh, you know,
00:36:59what portion is AI assisting humans, maybe 25%, what portion AI and humans collaborating deeply such that
00:37:07maybe 25 to 75% automation can happen. And then what portion can be completely automated without human
00:37:15oversight is what I think change management, you know, uh, uh, uh, is where change management is
00:37:21required so that leaders can think about how are my roles evolving, how are humans and AI collaborating,
00:37:28how knowledge gets, uh, uh, codified, how governance works, how trust is built. And I think trust is the
00:37:34most important thing here.
00:37:38Vernance was the second biggest result in the audience poll, which means it's a really important
00:37:47piece to delve into today. So what does AI governance look like in a regulated mortgage business where
00:37:55accuracy and audibility are critical? Siddhartha, you just touched on this. Let's delve into it more.
00:38:02Yeah, I think that, you know, um, with the advent of AI and AI, let's say doing a lot more
00:38:11work,
00:38:12you cannot have an opaque or a black box system making those decisions and affecting borrowers, compliance
00:38:19or investor risk. I think you eat the AI system needs to have explainability, uh, as to why did it
00:38:28make
00:38:28the decisions it did, auditability to be able to show, you know, if, if this is the, this is the
00:38:34piece
00:38:34of policy that was used and this is the loan data package that was used to come to this conclusion
00:38:39on
00:38:40whether this condition was, you know, failed or passed, et cetera. And one very important thing is it
00:38:45needs deterministic outcomes. Like today, LLMs, large language models and generative AI can be, uh, uh,
00:38:52uh, probabilistic. So the answers can change a mortgage. You cannot have answers changing based
00:38:59on, you know, something that AI did. So you have to use a combination of the generative AI and
00:39:05deterministic business processes. Um, then governance requires human accountability. You know, at the end
00:39:11of the day, if we ask ourselves, um, who owns the outcome, who's going to be held responsible for
00:39:17the outcome? I don't think AI is responsible for the outcome. It's humans that are going to be
00:39:23responsible for the outcomes. So what are the escalation paths? How are exceptions handled? What
00:39:28are the oversight mechanisms that are there? Um, and who, you know, is able to approve what? Those are
00:39:34things that we have to think about. And last is continuous monitoring. You know, if you think about,
00:39:39you know, when we used to take technology and move it into production in the past,
00:39:43yes, you did monitoring, but you didn't expect the answer to change. Well, now with something called
00:39:48model drift, where large language models can drift in production, you have to actually continuously
00:39:54monitor, uh, the outcome that is happening in production so that you can see whether, you know,
00:40:02we're still on track or not. And if we're not on track, then we have to be able to, you
00:40:07know, uh, make
00:40:08certain changes to check for accuracy, check for a drift, check for exception patterns, et cetera.
00:40:15And I think, you know, at the end of the day, people do not trust AI to do certain things.
00:40:20And
00:40:21I think this is all going to come down to how can the right governance and the right approaches
00:40:26create trust for executives, such as Eric, for the regulators in terms of how you can communicate
00:40:33as a lender to the regulator, that yes, this is, uh, uh, the right thing to employees, people who
00:40:38are using the systems, uh, to investors and then borrowers, you know, at the end of the day,
00:40:42they're the customers.
00:40:48Well, to quote the, the great Kent Brockman from the Simpsons, I for one, welcome our new AI
00:40:55overlords, and I'm happy to help them as they, uh, force us to toil away in their underground data
00:41:01centers. Um, you know, it's, it's really funny. I think we were talking about this before we, we,
00:41:06we came on, I've got three kids, a 15 year old, a 13 year old, a 10 year old, and
00:41:10the 13 and 15 year
00:41:11will feel very passionately about, um, things that they have very little knowledge about,
00:41:15which I think is a, is a pretty typical, um, uh, experience for kids that age. And,
00:41:20and, um, you know, lots of strong feelings about, about AI and there's, there's, there's power and
00:41:28there's danger in something that you can describe or put an identity around. And, and AI is sort of
00:41:33this, I think this vessel that people channel all of their anxieties into about technology, um,
00:41:40uh, the tail wagging the dog, not the other way around, you know? And so AI, from a governance
00:41:46standpoint, I think a lot of folks who run companies might be, if you're listening to this,
00:41:51probably know this. It's treated as this whole separate magisteria, this thing that we have to
00:41:56manage differently. AI is its own source of risk. And, you know, I think a little differently. I don't
00:42:01think the sources of risk that we're trying to manage have changed, right? You think about what we
00:42:05manage in the mortgage industry from a risk standpoint, data security, privacy, quality
00:42:10control, fair and unbiased lending practices, counterparty oversight, um, that those, those,
00:42:17those haven't changed. Um, we still have to manage them. And as we bring AI solutions into place, um, we,
00:42:26we, uh, regulators aren't good about considering counterfactual risk. Yes, there's risk when you bring AI
00:42:33tools in, but that that risk exists already. Are you making it better or worse? Um, is an AI engine
00:42:40going to make a biased decision? It's a risk you have to manage for, but, um, I guarantee you who's
00:42:47really good at making biased decisions is human beings. Um, we, we, we thrive at it. Um, and so,
00:42:53you know, how do you, some ways it's to, to the points that Arthur made, you've got 100% audibility,
00:42:58traceability. You can see the provenance of the decisions that are being made and why, um,
00:43:02you know, sometimes you can't always get that out of an underwriter, uh, hidden biases. They don't,
00:43:07uh, they don't even know that they have. So, uh, I think over time, my, my hope is that we
00:43:14move from
00:43:15a world where AI is seen as the separate thing you have to govern and rather it's, it's another,
00:43:21um, it's another use case of technology and you're managing risks that you're always managing and just
00:43:29thinking about how you need to adapt those oversight regimes for the specific, you know,
00:43:34way that AI uses it. So, um, and, you know, we're seeing that with regulators, uh, sitting here in,
00:43:40in Colorado, uh, a lot of news, Colorado was the first in the country to come out with a very,
00:43:46um, uh, a very strident AI policy. And, you know, they went through a couple of rounds, um,
00:43:52in the state legislature modifying those rules is I think, uh, even, even the, the, the representatives
00:43:59started to wrap their head around, okay, is this totally different or, or is this just an,
00:44:03the next technological horizon of risks that we are already asking people to manage? Um, and so,
00:44:11yeah, it'll take, I think it'll take time for companies, for people, for regulators to get their
00:44:16heads wrapped around what's, what's different and what's really just the same old risks wrapped in a
00:44:21kind of a new technology solution. Um, but, uh, you know, what we have to do as a mortgage company
00:44:28at the end of the day, isn't going to change. We, we get paid to, um, identify whether or not,
00:44:34um,
00:44:36money should be lent to somebody and at what price based on the risk. Um, and, uh, AI doesn't change
00:44:42that. It just gives us different tools for solving that question and then creates different, um,
00:44:47expectations for how we manage the risk. I think that's a fantastic point of view that I hadn't
00:44:56considered. So thank you, Eric, for that perspective. And before we wrap up with one of our
00:45:02last questions to Eric and Siddhartha, I have one more poll question for you guys and probably a very
00:45:10critical one, which is how far is your organization and adopting AI, where are you guys actually at?
00:45:17Are you just exploring? Are you running pilots? Is there a limited production use? Are you scaling
00:45:23across the organization or have you not started yet? Let us know where your organization is at.
00:45:37I have to wonder, Eric, if considering your last statement, if this is a similar
00:45:45and a similar inflection point for previous iterations of technological innovation,
00:45:50are there always the same concerns? Are these risks, similar risks to what we've been considering
00:45:57with the new technology wrapper? I think that that was a fantastic point to raise at this juncture. It
00:46:03doesn't have to be as scary as people might be thinking. The journal had a great piece on this
00:46:08a couple of weeks ago, um, talking about that using automobiles as a historical, um, example when
00:46:16that came, the disruption that, you know, automobiles had on society was probably even more profound than AI
00:46:21does is going to have here. And, um, you know, you could die getting thrown off a horse just like
00:46:28a car.
00:46:29You just needed better brakes on a car. So I, I, I, I, I think your, your push is the
00:46:34right one. This,
00:46:35this isn't that different than inflection points in the past. It's just, um, it's just powerful and
00:46:41it's, and it's, and it's, and it's here now. I was talking to someone who works on AI, uh, in
00:46:48the
00:46:48mortgage space and works on teaching people about AI at some of the MBA events. And her stance on this
00:46:56was
00:46:56that prior to the invention of airplanes, transcontinental travel seemed impossible
00:47:03for an overwhelming amount of the population. And now it's something that is entirely possible.
00:47:09And there's even airlines pushing to make it increasingly affordable. So perhaps what we
00:47:14think about as impossible now is similar to what people had thought was impossible 80 or a hundred
00:47:20years ago. Yeah. If you, if you think about it, the LLM model that you're using today, the AI you're
00:47:26using today, the worst AI that you'll ever use in your entire life, right? You can only, it's only
00:47:33getting better in a breakneck pace from here. That's a great point. So our audience results are in,
00:47:43we are very firmly in running pilots, which is fantastic to hear. So we have running pilots
00:47:50followed closely by just exploring, then scaling across the organization. I love that that is a
00:47:57close third and then we have limited production use. And then in a fantastic page of change of events,
00:48:04there is nobody who responded, not started. So love to see, love to see that, which is
00:48:11absolutely amazing. So Siddhartha, you just mentioned that this is the technology, the AI that we're using
00:48:18today is going to be the worst technology or the worst application of AI that we use. So if we
00:48:24have this
00:48:25conversation again in two to three years, what will be true about the lenders who leaned into AI
00:48:31transformation early versus the ones who waited?
00:48:37That's a very tough question. It's like saying, do we have a crystal ball or not? I think that
00:48:43those lenders who really think end to end about the transformation and the operating model change,
00:48:50which is all the way from a borrower to a loan officer, to a loan processor, to an underwriter,
00:48:55to a closer funder. And by the way, that's how we think about JazzX's solution,
00:48:59because the same work is being done to some extent by the loan officer and the underwriter and the
00:49:06loan process, et cetera. When they think about it end to end, I think they will be able to operate
00:49:12at a completely different level. They'll have materially higher productivity. They'll lower the
00:49:18cost, like the $11,000 cost of producing a baby versus $11,000 of producing a loan. They'll have
00:49:26faster cycle times, which means that the borrower experience is better and they're able to get a
00:49:31higher conversion rate from the borrowers into the funnel. And the organizations will be more
00:49:38scalable. There'll be fewer handoffs. So I think those who start now and actually focus on how they
00:49:45move that institutional knowledge out of the individual employees into the system of intelligence,
00:49:50I think they will experience a completely different economic base. Their top line will grow
00:49:58and their bottom line would improve. The cost of production would improve. So I think that
00:50:06that's how I see that those lenders who start now in two to three years will be at a completely
00:50:12different place in being able to manufacture that loan much faster and with a much lower cost base.
00:50:20Eric, I don't know if you, you know, would love to get your thoughts on that.
00:50:23No, I was just, I really liked the point you made. I never thought about this, that today's AI is
00:50:28the
00:50:28worst AI you'll ever be using. It reminds me of one of my favorite comedians, Mitch Hedberg, had a
00:50:34great joke where he says, I hate it when people come up and say, this is a picture of me
00:50:38when I was
00:50:38younger. And like every picture of you is a picture of you when you were younger. But I think,
00:50:48look, this is the future. I don't think anybody can avoid this. And if you don't adopt it,
00:50:53you'll get left on the side of the road. I'd maybe pivot the answer to what's going to be true
00:50:59about the,
00:51:01from a people standpoint at the companies that adopt it now versus those that adopt it later.
00:51:07We don't have crystal ball. We don't know which solutions are going to succeed, how fast it's
00:51:12going to go, what the savings are going to be, whether those savings will be offset by additional
00:51:16complications. But given that AI will be absolutely the way that we interact with our technology
00:51:25moving forward, I think those that start earlier will have an advantage in that they'll have more
00:51:32opportunities to have their team members learn how to use AI to become AI fluent. Honestly, it's one of
00:51:42the most terrifying things as a leader is new technologies that come up that you don't know.
00:51:50I mean, most people in mortgage got to a leadership position because they were the best at being a loan
00:51:54officer or understanding the technology. And it was mastery of the tools and the systems that got you
00:52:01to places of higher and higher authority. Now it comes in this disruption. And, you know,
00:52:05the 20 year olds that we're going to be hiring into our company, they are going to be native AI
00:52:09speakers.
00:52:12The rest of us, we're going to have to, we're going to be AI as a second language. And I
00:52:17think we're
00:52:18always going to have an accent. And so I think the more, the sooner we can train and give our
00:52:25team
00:52:25members, even those that might be, you know, 30 years into their career in this space, opportunities
00:52:30to learn how they can use AI and how they can actually become fluent in AI. Those workforces are going
00:52:38to be better positioned to pick up quickly on whatever next iteration of technology comes to
00:52:43the fore. So no different than 40 years ago, 50 years ago, companies that taught their employees
00:52:49how to use computers and gave folks an opportunity to, you know, engage with, you know, digital technology
00:52:56were the ones that could move quickly as that technology continued to accelerate at a, you know,
00:53:02an exponential pace. Same thing with AI. Any, any mortgage lender that, you know,
00:53:07a year from now can't point to at least one use case where every one of their employees is using
00:53:15AI
00:53:15in a meaningful way. I, I think you're, you're putting yourself at risk.
00:53:20Well, and if I, if Eric, if I can add to that, you know, the people part of it, two
00:53:24things. One,
00:53:26you know, in a recent conversation, the VP of underwriting actually got tremendously excited
00:53:34because she saw the ability of being able to change the process herself because it could be done through
00:53:41a natural language interface yet governed by the system rather than having to figure out, okay,
00:53:47who do I go to in IT? How long is this going to take? What are the IT resources? What
00:53:51are the IT
00:53:51skills that could be taken? And then she said in a lot of cases, she just didn't do the change
00:53:56because it was
00:53:57not big enough to have all of those needs. But now she's like, I can actually streamline my processes
00:54:04myself. I'm in control of my own destiny. And one of the tenants that I have set for my organization
00:54:11is, um, you know, you, uh, AI first innovation, which is refuse manual toil, ask yourself,
00:54:19where can I use AI? And why, why can I, you know, and, and really try and answer the question,
00:54:25why can AI not help me here? But then if all you do is give us the answers from AI,
00:54:31then we don't need you. So then you have to think about, well, where do I bring in judgment?
00:54:35Where do I bring in my knowledge and my experience? And how do I make the answer or the, whatever
00:54:42output I'm providing, whatever outcome I'm delivering completely different and much more
00:54:47powerful where the manual toil is gone, but it is so much better and done so much faster.
00:54:56Absolutely.
00:54:59There, uh, there's a, Joe, I think that does mean you might have to give up your calculator. I'm
00:55:05impressed that you still have an HP. I do think it's also interesting where this probably is similar
00:55:16to where you, we spoke the example of using a computer training on a computer came up,
00:55:22but also probably the transition from paperless to digital. There's probably a,
00:55:29a process to get used to all of those systems as well.
00:55:32And then the good news is that everyone has gotten used to it. Meaning who hasn't used a Gemini,
00:55:38I'm a Google fan. So who hasn't used Gemini or who hasn't used chat GPT or who hasn't used
00:55:44Anthropic Cloud, right? Everyone's used it in some way, shape or form. So the great thing is
00:55:49people can see the power, people can see what it can do. But I think people are also realizing where
00:55:55they need to bring in their judgment, right? Uh, for example, students, when they're working on their,
00:56:00uh, papers, you know, one of the professors said, look, the HBS case studies that you need to analyze,
00:56:06feel free to go ahead and, you know, use AI. You can get the answer to it. But what I'm
00:56:11going to do
00:56:12is I'm going to have you have these personas that you have to interact with that represent people in
00:56:17that case study. And now you've got to go figure out how to engage with them, how to partner with
00:56:22them,
00:56:23but delivered through AI, right? Those personas are AI personas that are simulating that Harvard
00:56:28business case study. So it's not enough to just get the answer. So you have to think differently
00:56:32about how you do that.
00:56:35All right. Well, we are rapidly approaching the top of the hour, uh, the top of the hour. Eric,
00:56:41Siddhartha, thank you so much for sharing insights with us today, for giving us multiple points on
00:56:47governance on how to get started on the balance between AI and legacy systems. I want to turn
00:56:54things over to you guys to give final thoughts before we wrap up. Eric?
00:57:02Oh, uh, nothing. Just, um, I don't have any deep thoughts. Uh, I guess just more, um, I think
00:57:10it's easy with this new technology to, um, feel like you're, you're behind. I love that question
00:57:18you asked earlier, you know, anybody not doing anything, I can't imagine anyone in their right
00:57:23mind admitting they're not doing anything, even if they're not doing anything right now. Everybody
00:57:27feels this, this anxiety, like I'm supposed to be doing more. I, I, as leaders, I'd say don't
00:57:33abandon your judgment. Um, this is a new technology. Um, uh, take a measured approach,
00:57:41find things that work. The opportunities will present themselves for when you need to hit the
00:57:46gas pedal, but this isn't like, once again, this isn't the first time those opportunities
00:57:51haven't presented themselves or you've had to navigate new technology or challenges. So I'd say,
00:57:56you know, uh, just for any, uh, folks who are, uh, similarly wrestling with how do I play? Where do
00:58:03I
00:58:03play? You know, uh, when do I jump in? Um, this isn't different than any other decisions you're
00:58:09making. So trust your judgment and, um, you know, uh, be smart about where you, where you deploy resources
00:58:15and you're probably not as far behind as you think you are. Yeah. I w I would just say, you
00:58:21know,
00:58:21if there are two takeaways from here that, you know, that I also heard from Eric is one, um,
00:58:28think end to end about the operational transformation that this is going to drive.
00:58:34Don't just think of this as just a piece of technology in a particular silo that you cannot
00:58:38automate, but take the leap to think end to end around how you can transform your entire mortgage
00:58:45production. And the second thing is think about how you can leverage AI to institutionalize
00:58:51the judgment that today resides in the minds of the folks, the quartet that Eric was talking
00:59:00about. How do you institutionalize that judgment such that you can do better reasoning, better
00:59:05automation? And lastly, you know, what is the role that humans should play and where around judgment?
00:59:14Those are three things. I guess I said two, but those are three things that I would, you know,
00:59:17uh, uh, uh, you know, say would be takeaways and Alison, thank you so much for being a wonderful
00:59:23moderator. Really appreciated all your questions and the polls, reaching out the polls and helping us
00:59:28understand where those polls landed. That was wonderful. Thank you both. I have several pages of notes,
00:59:35so I appreciate both of your perspective and to our audience. Thank you for joining us. That wraps today's
00:59:42webinar, rewiring mortgage for the AI era. Why AI is now a leadership imperative. Eric Siddhartha,
00:59:50once again, thank you so much for taking the time and sharing your expertise and our audience. Thank you
00:59:55for joining us. Housing Wire will be sending a recording out of today's session to everybody who
01:00:00registered and it will also be available on our website. Thank you for being with us today. We'll see you
01:00:06all again soon.
Comments