[David] Listeners, we’re back with Garrett Locklear, the creator of Candid, an amazing technology. And he has a new announcement or another announcement of developments he has going. Garrett, good to have you back on the podcast, friend.
[Garrett] Thanks for having me, David.
[David] Tell us the big announcement.
[Garrett] Well, David, we’ve been working behind the scenes on something that is just truly amazing and first time ever announcing it in public, we’ll be launching in the next few weeks our AI infrastructure cloud to allow our customers who are mortgage organizations in the marketing and sales world not only to take advantage of all these exciting new features in the world of AI. But do it compliantly, do it securely, and most importantly, do it natively inside of our operating system.
[David] So when you say natively, for some people hear that and they go, Like, yeah, kinda understand that, maybe I don’t. Explain natively.
[Garrett] Well it’s always been the vision of Candid not to go out and API into anything that you need, but instead create native experiences inside of the walls of our operating system so that you don’t have to worry about API and compliance issues and maintenance issues and management of vendors and things like that. And because you bring things in natively, you can also create a better borrower experience at the end of the day, where the borrower goes to one place for all mortgage needs from the very first time you speak with them all the way until years and years after closing, as they’re maybe looking at a home buyer report or completing an annual mortgage review. Same ecosystem, same operating system, one experience and a better experience, I would argue, for home buyers.
[David] Yeah. Well, let’s get into talking about some of the new features and benefits that you have developed. I got a chance, listeners, to look at these yesterday. And I gotta tell you, I’m just amazed at where AI is going with the tools we have, and specifically what Garrett has now developed and vetted, and it’s available for you. So walk us through it, Garrett.
[Garrett] Yeah, as part of our AI cloud infrastructure, the first thing that we really thought about was and were hearing from our customers is they’re constantly being asked by loan officers, hey, do you have this AI tool? Do you have this AI tool? And the pace of change, as you know, is so great, it’s just hard to keep up. And so the one thing that we wanted to address in that situation is you should not have to wait on a vendor to roll out an AI tool that may or may not actually be what your loan officers are asking for. Instead, our customers should have the power to build these AI tools themselves. And that alone sounds pretty scary. How in the world am I gonna build an AI agent? Yeah.
[David] Yeah, letting that gets you created the wild, wild west and then you talked about it being compliant. How in the heck do you let them develop something and it be compliant? Gotta put some construct around that for us.
[Garrett] Yeah. Or even how do you just how do you build it? Like some a lot of people are they can’t answer the question, what is an AI agent? How do you build one if you don’t know what it is? And so we th we thought yeah.
[David] Yeah, let’s go to something else. Let me go to something else just before that. I always like trying to understand the problem. So with what you’re about to talk about, what is the problem that you’re solving?
[Garrett] The problem that we’re addressing with what we’re calling our AI agent studio is when that loan officer comes to their corporate counterpart and one of their colleagues and says, Hey, I really like this tool over here, I’d really like to try it out. But mortgage organization looks at it and like, well, is it compliant? Where does the data go? Does it train on the data? Does it leak the data? There’s a lot of questions around AI and how secure it is. And the spectrum right now we’re hearing is some mortgage companies are shutting it down completely. If it has the letters AI in front of it, we don’t allow it. And then other mortgage companies are on the on the opposite end of that, allowing the loan officers to go out there and try these tools, but then that runs a huge risk. And so when we look at that, what we develop as part of our AI infrastructure is we’re calling it AI Agent Studio. And it is a no-code, not even drag and drop studio natively built into our infrastructure where through natural language prompts, mortgage companies and the admins there can go in and build what the loan officers are asking for. So I’ll give you a simple example that we’ve seen. The loan officer came to their colleague and said, Hey, I really would like to be able to automatically create referrals that drop into my email inbox. And that’s beyond automation because you have to look at an email and understand the intent of Right. And so there’s some companies or vendors out there do it, but then you have to worry about the integration and compliance and security and all that. And so what happened was d the email that the loan officer sent copy and pasted into the AI agent studio prompt. And then the AI agent studio went to work building exactly what the loan officer was asking for. And there’s some back and forth, but it’s like you and I having a conversation. and so the AI agent studio builder will come back and it’ll architect what it thinks that you need and it’ll say, is this what you’re thinking? And it may be exactly on, it may be a little bit off. But then you have a response back, you say, this is perfect, except this one thing over here. And then it goes back to work building and crafting this AI agent that the one officer…
[David] That’s fascinating. So walk us through how this is done.
[Garrett] Well is so the AI infrastructure cloud and specifically the AI Studio, it’s using a large language models on the back end to be able to understand your intent and your goals and what you want to build. And then behind the scenes, it’s actually building the code for you. So that you or I or anyone doesn’t have to learn how to code or even learn how to drag and drop, which is kind of indicative like an automation builder. Instead, all we have to do is learn how to talk to it. And that that bleeds into the second thing that we’re rolling out, which is also part of the AI infrastructure. We’re calling it Candid Coworker. And this to me is just wild, David. because you know you know my story. I in about eight years span of time, I grew my personal production from eight million a year to two hundred and fifty million a year. But it took eight years, right? It’s a long time to be dedicated to your craft and to grow it and to address all those issues that come along. I see AI is taking that eight years and shrinking it down to two or three. That’s how powerful I think AI is getting. And with Candid Co-Worker, this is a fantastic example of that. So think of Candid Co-Worker, which is a second release that we’re doing within our AI infrastructure, as a loan officer’s personal assistant available 24/7. And instead of let’s say the loan officer wants to figure out who they need to call, right? Maybe it’s a lead, maybe it’s a realtor, whomever. Instead of going to a report or a dashboard or maybe scrolling through a list view, instead, I would just ask candid coworker the question. And I would say something like, hey, I want to know the top 10 lists of realtors by their volume who I’ve worked with in the past but have not sent me a referral in the last 45 days. Right? and for me, back in the day, like I’d have to go through spreadsheets and then I might have to build a report, but it would be you know a solid hour of research in this. And by the time I’m done, I’m already like a little tired and I don’t even want to call them anymore. But here I can get to that information within about 30 seconds. And on top of that, because we have AI insights, it’s already in the platform, but it’s pulling on all this data for realagers and and other parties, because we have that. And
now we’re launching Candid Coworker. I not only can figure out who I need to talk to, it will tell me how I need to talk to them based on their personality. And then should I send a text? Should I send an email? Hey, create a plan for me the next 10 days, the next 30 days on how to engage each of these 10 realtors that the system has identified. and by the way, just go ahead and schedule it out for me and do it. And that might look like…
[David] I watched you do this yesterday when you were I was literally we’re doing a screen sharing session and I watched you write a command. If you could walk our loan officers through that command or what you wrote, I mean again, it just texting, or you could just speak it in. In your case, you were typing and walk them through it and and then how you went back and said, yeah, but add this and add that. And it’s it’s no different than some of the large language models. So walk walk me through because it was so fun to see what it did. Almost want to go back and be a loan officer again.
[Garrett] That’s what I’m saying. It’s going to shorten the time, the distance between zero and fifty million, zero and a hundred million, because you can do things so much more faster, so much more efficiently, and you’re not shooting in the dark anymore. Right? Our our the Candid coworker is layered with, in this example, it’s layered with realtor production data, realtor personality data, realtor social data. You no longer have to guess anymore. And on top of that, it’s not just about finding, you know, who’s the who’s the realtor who does the most volume. Right. Even that was like pretty cool and new up to about, you know, five or six years ago. Now it’s about it’s not just about the volume. I want you to find me the realtor that you think I would work best with based on the last hundred loans I’ve closed and the realtors tied to those. Right? So it’s looking at your data, creating profiles based on their personality, based on maybe the part of town they work in, maybe the purchase price range they work in, maybe the age of borrowers. It’s creating a profile of a realtor, and then it’s going out and saying, Hey, I think this realtor over here is a pretty close match. Would you like me to write an engagement script for you? Absolutely wild stuff. But because it goes back to the heart of CANDID. Because we have all the data in one place, we have immense capabilities. Because we have all the native functionality in one place, we have these capabilities. We have the capability to send the SMS and the group SMS and the video SMS or the video email or the handwritten card or the voicemail drop. We have all these capabilities, all native, all unified under one operating system. So now we layer AI with it and the opportunities are extraordinary.
[David] Yeah. And I I mean I sat and watched it, folks. It was just amazing. So explain for those, again, we’re not able to do screen sharing. Well, we could, but most people are listening to this while they’re working out or jogging or driving the car. so describe what that brought back as a result. That was so impressive.
[Garrett] Yeah, well I think so when we talked yesterday, there’s about two hours left in the day. And so one thing I just put in there is as I said, hey, I got two hours left in the day. I want you to tell me what’s the absolute best use of those last two hours based on look at my prospects, look at my realtors, look at everything that’s going on, my incoming text messages, my missed voicemails, all of it. I want you to tell me what’s the best use of my time. And I pulled it up. It gave me the two hour plan to end my day and one and a couple of things were like, hey, you’ve got a missed phone call. Here’s your voicemail. You should definitely return this because this is a borrower. They’re looking for a $2.5 million home. It had the transcripts, right, from the voicemail. I had some a couple of unread text messages. So same thing. And then it identified, hey, here’s three realtors that you should reach out to by the end of the day. You did a lot of business with them in the past. You haven’t done a lot of business with them recently. And then by the way, here’s the top three prospects you should touch on before the end of the day because they’re showing high probability of buying. But it was able to regurgitate all of that to me in about 30 minutes because it has access to the data and it understands the data structure. It understands who I am as a loan officer in that scenario. But all we’ve been talking about so far are loan officers. Who else uses our platform? We have branch managers, we have regionals, we have people in recruiting departments, we have people in marketing departments, we have admins. And so CANDID coworker doesn’t stop at the loan officer. It can touch every one of those roles within a mortgage organization. So think about the branch manager who needs to know who’s the top five and the bottom five producing loan officers in their branch. Now you can surface that pretty quickly, which again, like that’s a reporter dashboard as well, you know, big deal. But you can say, hey, for the bottom five, I want you to create a 30-day plan to increase their production based on the attributes you’re seeing with my top five loan officers. What are those top five loan officers doing that I can apply to these bottom five? Right? And boom, instantly now you have a 30-day plan that you can walk through with your bottom producing loan officers to bring up their production. Yeah. Or
[David] Wow.
[Garrett] Maybe it’s as simple as like, hey, you have 15 voicemails and 13 unread texts. It could be something even more simple. I want you, I want you to tell me why they’re the bottom five producing loan officers versus the top five. And it’ll go in and look at the data and help you understand what’s going on. It could just be a lead funnel issue, it could be a lack of leads, but it’d also just be a conversion issue, not necessarily something wrong with the loan officer. But how in the world would you know any of this until something like Candid Coworker comes along and helps you dive into the data, surface it very quickly and then create plans for you.
[David] If you’re managing people or you’re managing a pipeline, a relationship with real, it’s turning out that the new skill set that’s really going to be accelerating someone into success is learning how to ask the right prompts to get the play.
[Garrett] Yep, asking the right question.
[David] Garrett, the biggest challenge when you have tools like this is learning how to use them. And what we’re doing is we’re finding with LLMs and what you’re built now into CANDID. There’s a new skill set or a skill set of being able to ask the right questions. How would you help people get there? I want to have a discussion around that because I think as important it is to talk about your technology. It’s more important, how do you use it?
[Garrett] 100%. Start simple is my first my first advice. so you know we’ve been talking about some prompts that like include two or three different data points, right? Like cross-reference this and do that. But start simple. Do I have any unread text messages today? And then it’ll say, Yeah, you’ve got you know two or three. And then you could follow up with like, all right, well, what would you recommend for response? Right? Like start really simple. You don’t have to be complicated with it. Talk to CANDID coworker like you’re talking to anyone else, like you’re talking to a loan officer or assistant, and keep things simple and then it’ll respond, you know, with the information you’re looking for. But then the cool thing is underneath it, it’ll also say, Would you like me to write a re-engagement script for you based on the personality type? Right? It’ll actually prompt you beyond a simple question to think, wow like, wow, I didn’t know you could do that. Yeah, that’d be great. Or, you know, I’m thinking about email. Could you write an email instead of that? Right, but it’s just having a conversation back and forth. But the best way to start learning how to get really good output is to focus on quality input. And quality doesn’t necessarily mean quantity, it just means quality and you can keep that simple.
[David] Yeah, well, it was so fun to watch the output that when you created the prompt and then made a few tweaks to that prompt as in the iterations as you were showing me yesterday, then it spit out an exact time by almost minute by minute what you should be doing, how you should do it, and cut and paste this into this email or this text, do this. I mean, it’s like having a coach right there that has having a more
[Garrett] Yeah, having a mortgage coach combined with a mortgage coach who’s intimately has intimate insight into all of your data, which no one person could, and then can answer based on all of that data. I had a mortgage coach for about a year when I was ramping up my personal production. fantastic, but he didn’t have access to layers and layers and layers and layers of my data. And so the advice that I was getting from the mortgage coach was good. But it could have been better if he had known all the data. How many clients have looked at a home value report, completed their annual mortgage review, completed a mortgage application? How many clients have you called? How many of those did you answer? What’s the transcripts? And then being able to analyze all that in about 30 seconds, impossible for a human, but for a large language model embedded natively into Candid, very simple. And that’s loan officer. We talked about branch manager, but think about recruiters. We have a recruit cloud. And because we’re layering in nationwide data, loan officer and realtor data, the recruiter can do the exact same thing. They can surface information, they can create engagement scripts, they can plan their day. Same thing that a loan officer would do just from the recruiting angle. It’s just absolutely wild because I was talking to a company yesterday and I was telling them most people just get like that top 100 list, right? Top 100 realtors in your area, top 100 loan officers in your area. And that’s fine to work through. But a lot of times you won’t find long-term success with that. Even if you get a win, it might not be a great fit for the company. And so they get in there, they produce a few loans, and things are just rough. And they eventually turn a year or two later. But as a recruiter, you can go even deeper than that, look at not just their production data, but who are they working with? Is a realtor? Do you run in the same quest circles? Is their personality type a good match? what type of volume are they doing? And so you can create a profile internally, because you have the data, around your top producing loan officers. And then you can take that profile and you can say, hey, I want you to go now out to the market and I want you to find loan officers who match that profile. And in a really perfect world, you would find a loan officer doing 10 million, maybe 15 million a year. And you’d bring them into the organization and then watch them just scale and grow to a hundred million dollar producer because you’re able to identify the type of person that fits well with your organization, not just their volume.
[David] That it gets so exciting. We could go on and on about this, but that’s not the only you’re releasing. You’re doing more.
[Garrett] Well, we’ve got as part of the AI infrastructure, we’ve got Candid Coworker, which I think truly will change the way our users interact with our platform. And the best part of it is you don’t even have to be inside of candid to use it. Candid Coworker will work inside of Teams, it’ll work inside of Slack, it’ll work, of course, on the mobile device. You can install it in a lot of popular systems out there so that you can provide this incredible tool that has access to all this data, no matter where the loan officer is working. But in addition to that, we’re releasing a few custom AI agents as well that are needed to the platform. One’s a business development manager that sits on the account side of things, to think your professional relationships like realtors. And it’ll do autonomously what we’ve been talking about a lot, quite honestly. And it’ll go out and it’ll find the best fits for loan officers in their area that they’re not working with. And then it’ll engage in a value-based type of marketing way. It’ll engage these realtors autonomously for the users and be able to offer them a different variety of different value ads. Maybe it’s a property site for the upcoming listing. Maybe it picked up on social media that they are really frustrated with their current lender. And so it’s just a good time to reach out to them. Whatever it looks like, it can it’s it has the capability to do this autonomously for the loan officer as the loan officer is already working with their clients and referral partners. And so back to what I am almost claiming is a prediction. I do see loan officers going from zero to a hundred million in less time than they ever have because of these powerful tools. And I also see the average production for loan officers. Of course, it’ll continue to go up just like inflation and depreciation of home values do. But I think it’ll outpace that 2, 3, 4 X because the loan officer’s day is now focused on relationships and not all this busy work. What does that mean for their volume of per loan officer volume? I see the average per loan officer volume just skyrocketing.
[David] Yeah. We’ve been talking about that for a while. I say s look at tools like Candid to accelerate, expand your business. I think it’s distinctly possible and it sounds ridiculous, but it’s now s I see how it is. And it’s you can ten X your business. I mean it’s
[Garrett] Yeah, I do I do believe that. I don’t think that’s hyperbole.
[David] Well, and not only that, yeah, look at yourself, you are doing that. You look at how many you produce a significant volume. I can know if you want to share that, but you produce a significant volume and the number of hours that you’re working on that, you have a you have a tech company and you’re a top, top producer for one of the leading mortgage companies in the nation. And that’s not possible. And we’re and have a home life. You’ve got a folks, if you can see his family, it’s a beautiful family. It’s one of those, it’s one of those postcard families. why’s beautiful? The kids are beautiful. They’re just I know you’re not. We’re gonna throw a picture up, but yeah. I don’t
[Garrett] Stop it. As long as I’m not in the picture, David. I don’t wanna make the rest of look.
[David] Well, we don’t we often times what the milkman looks like, but that back in the day things okay. But the truth of the matter is you have time for work-life balance too, you’re it. And I think this is to be able to do those kind of things. We’re not talking about people killing themselves or losing their life balance or not being able to have other interests. You certainly do. You have a tech company, you’re a top producer, and a beautiful family that You spend time with.
[Garrett] Yeah, quality of life is super important. And a lot of loan officers will have four closings a month, five closings a month, and work sixty hours a week. And I’m they don’t believe me what I’m telling That’s that’s crazy. I I work a a fraction of that and close a hundred million a year. But it’s one hundred percent possible if you have the systems and the process and the people in place to do it. A hundred percent possible two things I wanted to mention as part of our AI infrastructure rollout. So we’ve got the studio builder, we’ve got coworker, we have our business development agent, we also have a processing agent and a pipeline AI agent now, all natively built into the platform. And when you think about from the perspective of from the perspective of a loan officer, what takes a lot of their time when they have a new client come in the door, it’s reviewing documents. And so what we’ve done is we’ve built an AI agent that exists again natively into the platform, behind our trust layer, 100% secure. And it evaluates and has real-time back and forth with borrowers as they’re uploading a driver’s license or a bike statement or whatever it is. And there’s things that just have to be done every time that right now create a bottleneck, a human bottleneck. So when that borrower uploads a driver’s license and it’s expired. Well, it just sits there until a loan officer or LOA looks at it and hopefully cashes as expired. But why in the world do we not just let the borrower know that in real time so they can upload one that’s not? Or that last page of the bank statement that’s missing, or that entire PDF that’s password protected. Stuff like that. Like we’ve got to get the basics down. And once you sanitize the documents, then they drop in and the bank statement. A loan officer should no longer have to look through pages of a bank statement. If there’s something to look through, then surface it and let them know. If not, then clear it. That’s what our AI agent does. But then it goes beyond just sort of the back and forth with these documents. It reaches outside of them as well. So, example, purchase contract comes in, drop the purchase contract in. Well, we need to know the phone number and the email address and who the listing agent is, along with title. You probably already know the buyer’s agent. But a lot of this information isn’t even on the purchase contract. So we’re not only extracting data out of these documents, but we’re also going and finding the data that’s missing and populating that as well. And so because our system already has the capability of order auto ordering title insurance and homeowners insurance, the documents actually become the command line that drives the next action, which we call documents as a command. And so you can get a purchase contract, and that purchase contract can dictate ultimately that a title order needs to be placed. And a human wasn’t involved in any of that. So you’ve taken like a solid 45 minutes, I would say. This might be on the low side too, but you’ve taken a solid 45 minutes that takes a loan officer to work through all of these documents, and you dwindle it down to under five, assuming anything needs to be looked at. But if it’s a pretty clean file, they might not look at it at all.
[David] The world is changing. What’s your prediction as you continue to develop Canada? Where any idea anything you can tell us of where you see this all?
[Garrett]I wish I had candid 10 years ago when I grew my business at 250 million a year, but I didn’t. I my prediction with where things are going is it’s never been a better time to be a loan originator. I think you need two things and really two things only. You need to be an expert at relationships and you need to be an expert operator. Everything else is just noise. If you’re really good with relationships and you’re an expert operator with your technology, then that I think that the sky is the ceiling, especially when you have a refi wave. You’re going to see loan officers produce more volume than any single loan officer has on on planet Earth ever. And you’re going to see the amount of time that it takes a loan officer per file continue to go down and down.
[David] Interested. Hence more volume. Garrett, so exciting. Congratulations on your continued success, what you’re developing. I love the way your brain works. Sometimes I want to get you on and just talk about some of the life hacks you do to enjoy life at another level. I mean, I’m thinking about how you travel at times. So folks, you known by the company you keep. And I enjoy Garrett at a personal level and a professional level, but then part of it is just. Had never thought of doing this. I could I could do these different things to enhance your life. And so one of the times I want to talk about how you travel because how you roll is pretty cool. It’s just and…
[Garrett] I appreciate that, David.
[David] Yeah, it and what’s so fun about that is opening and expanding your mind, people’s minds to what is possible today with
[Garrett] Yeah. That’s it’s super exciting. I couldn’t agree more. And David, I want to take a second to say thank you to you for being a mortgage industry leader for longer than you care to admit. I appreciate all that you’ve done for our industry.
[David] Well, thank you very much. It’s time for you to say, I’m s just turned 76.
[Garrett] I wanted to throw a number out there. I didn’t do it, David. Didn’t do it.
[David] Garrett, thanks so much for joining us. I appreciate you being here. Have yourself a wonderful rest of your day. How can people learn more about Candid and you?
[Garrett] Shoot me an email, Garrett@candid.inc, or come visit us online at candid.inc.
[David] Perfect. Garrett, thank you so much for being here, friend.
[Garrett] Thanks, David.
[David] Appreciate you.
Important Links:
I have personally originated over $1 billion in mortgage volume and founded CANDID to revolutionize mortgage marketing and sales experiences. Instead of relying on outdated spreadsheets and fragmented tech systems, CANDID uses an open architecture technology to consolidate and enhance mortgage marketing and sales experiences. It power the Modern Mortgage Experience for loan originators to grow their business, boost revenue, and free up time for focusing on strategic, big-picture goals.
I am a motivating leader with extensive expertise in the financial services industry and technology, particularly in how it can enhance relationships. I prioritize serving others and strive to make their experience with CANDID a standout aspect of their work. I guide my team to focus on fostering trusted professional relationships that deliver mutually beneficial outcomes.