EP04: AI Meets MLS: Turning Knowledge Into Answers

Episode Transcript

 

Shannon Baird  00:00

I’ve seen issues from outdated information that’s on public facing sites that have never been removed, even though they’ve been updated.

 

Justin Lundy  00:09

We built Navigator, which is essentially an AI assistant that learns on all of your documents, but it doesn’t it doesn’t take months to train.

 

Katie Smithson  00:18

Navigator does not hallucinate answers. It is going purely on the documentation provided by the MLS or the association. Technology is moving so fast these days. And yeah, I don’t think it’s a matter of like years that we’re talking about. It’s much shorter. What does the future of AI in the MLS world look like. Welcome to REpod — the organized real estate podcast, for the industry, by the industry. This is where MLS leaders, association executives, and brokers come to rethink operations, simplify complexity, and build systems that actually work together. Join the REcore team as we challenge outdated processes, explore smarter technology, and have honest conversations about the future of real estate. Let’s get into it. Welcome to another episode of REpod. Today. I have two very special guests. I have, Justin Lundy and Shannon Baird, both from Lundy. So want to take a moment to welcome you both. Thank you for being here today.

 

Shannon Baird  01:26

Thank you. I

 

Katie Smithson  01:27

want to just start off. For those who aren’t familiar with Lundy, Shannon or Justin, feel free to take this. But why did you start Lundy?

 

Justin Lundy  01:35

So I started Lundy a number of years ago with no idea what it would turn into. I actually was just trying to solve a problem that was close to me. My my mother in law, started losing her vision. She has what’s called RP, which is retinitis pigmentosa, and essentially the peripheral vision starts vanishing, until you eventually go blind. And you know, at the time, I was actually selling houses as a realtor. And our thing, Diane is her name. She used to just text me photos or really just listings on Zillow all the time, not anywhere in our market, just like, Oh my God, look at this house, right? Like that was our thing. And, you know, as as she started to lose her vision, and we were exploring all the resources available to her, you know, for everything, how’s she going to get around? Should she get a dog, or should she get a cane? Is she going to need a driver? Like, you know, everything, we realized that the internet is something that for her, it’s just going to be gone. And so, you know, with the internet, go search. So really, that. The idea was, how do we give her an audio version of search? And as we started unpacking that and building what’s now called finding homes, we realized that we stumbled upon an area that had been totally overlooked, like in today’s age, there’s still no real solutions. And so that’s, that was the origin of it. We, you know, accidentally solved this problem that I thought only one person had. And it turned out it was, it was pretty significant that brought us, you know, fast forward to today. We, we spent a long time like before, before llms were let out of the lab. Me and Shannon were staring at spreadsheets trying to turn real estate data into into conversation. It was gnarly, like one of our first deals. I just got LASIK done, and I was so worried that we were gonna it was gonna take us so long to launch this product that I didn’t care that they were like, don’t look at a computer for like, two weeks. I was staring at spreadsheets with sunglasses on trying to, like, turn real estate data into, like, this thing that you could talk to. So, you know, fast forward a couple years. We started scaling that, you know, dozens and dozens of MLS across the country saw, you know, the problem. They they they jumped in and did their part to adopt finding homes. And by the time llms were introduced, we saw instantly that this was the thing that finally unlocked voice technology, and we were probably one of the first to take it and sprint with it like we threw away all that code that we had written and swapped out Alexa’s brain for an LLM and that that gave us the experience to take that advanced AI and start using it to solve other problems inside of the industry. And you know, voice unlocks a lot of efficiencies, right? And large, language models solve a lot of problems that the industry was struggling with, and still is so long. Answer your question, you know, what is Lundy? It’s now, it’s an advanced AI company that that harnesses these tools. To solve all the problems that we see, and our customers now start bringing to us and you know,

 

Katie Smithson  05:07

well, yeah, I think that’s that’s a great segue to something I wanted to talk to Shannon a little bit more about. Is Shannon, obviously, you do a lot of onboarding with MLS customers. You’re talking to a lot of leaders out there, what are some of the like, common bottlenecks that they come across, some of the issues that you all have helped solve by with, you know, with some of the products and services that you’re offering?

 

Shannon Baird  05:34

Yeah, so obviously it started with making data more accessible, specifically to the visually impaired community, and that has certainly morphed into a lot of different potential problems and actual problems that MLS is an association. See, I think, in terms of the bottlenecks, it really depends on the MLS. I’ve seen issues from outdated information that’s on public facing sites that have never been removed, even though they’ve been updated, and to compliance teams coming to us with, like, running out of five plus rule books, to complications from, like migrating from information from one platform to the next, and that has, of course, there gets to the issue of Like, having accurate information, one source of truth, and putting that out to their customers in order to reduce confusion on to the agents and to the to their subscribers. And so that’s sort of what led us to Navigator and building that tool specifically in order to alleviate that issue, and having a central repository of information that MLSs and associations could use and to not have to worry about all the things that are swirling around in the ether, and that was why we built it in the first place.

 

Katie Smithson  06:54

Yeah. I mean, I know, I kind of know the story of how Navigator came about, but I guess, like, what was the the path to get there? I mean, obviously you had finding homes. And I don’t remember the exact timeline, because I know you also have the listing and the voice listing input, where did Navigator kind of come in, and how did, how did that idea come about, come to fruition? And then I don’t know just a little bit about that journey and how it got from that initial idea to where it is today.

 

Justin Lundy  07:26

Sure, I think it was a similar, similar timing to listing input, but, yeah, actually, it came out of a request Navigator did. I was at a dinner with Justin Landon, and we, he was like, Look, this is the problem, right? We’ve get, I mean, in their case, sometimes 1000 calls a day, and they live out of 1000s and 1000s of pages and documents. There’s rules, there’s compliance, there’s NAR stuff, there’s their association, specifically, there’s their MLS, because netris is a wholesale model, so they don’t even handle support. And you know, the common complaint amongst their members was like, I could call three times with the same question and get three different answers. Yeah, right. And it was like, he’s like, can you solve this? And it was a fun problem to think about, right and back then, nobody had harnessed the LLM in a way that that was solving that problem. So it was fun to tackle. And so we built Navigator, which is essentially an AI assistant that learns on all of your documents, but it doesn’t it doesn’t take months to train. It’s instant. And originally, it was built to augment your staff. Just give them superpowers, right? Instead of them having to put you on hold and look through all the different documents that may contain the answer to your question, you just ask, right? That they can just ask the Navigator in their little dashboard, and it gives them the correct answer. It gives them verification of the answer through citations, and they can just answer right and on to the next problem. That’s how it started. And then it’s yeah, actually, with a partnership with California Regional and your are their amazing support team, we we gave it to them as like a Will you try this and see if you like it. And they really just,

 

Katie Smithson  09:23

they’re like, We like it, and we have lots of ideas, right?

 

Justin Lundy  09:26

Could like it. No, they like they liked the foundation, but for it to be useful to them, and then, you know, translates to all of the support teams around the country. They needed it to be modified a bit, and so they gave us all their feedback, and it was amazing. So we, we took that feedback and implemented all of it, I would say, even the UI, you know, everything, and turned it into a product that that massive. Team can use to support their huge market, which means, you know, it’ll work for any size market. So it’s really kind of morphed into, you know, a staff augmentation tool to anything you need it for as as an organization.

 

Katie Smithson  10:18

Yeah, and I think that’s another interesting point, because Shannon has done a really great job of, like, putting together different use cases. So Shannon, I don’t know if you can, like, pull something out of the top, off the top of your head, but just like, a random use case that somebody has come to you with that you’re like, Huh? I never thought of it that way, but you could do it like, through Navigator.

 

Shannon Baird  10:41

I don’t know if it was one that was random specifically, but I think one of my favorite sort of stories that I heard was, we had a smaller MLS. They wanted Navigator. They weren’t sure. Like, they had such a small support staff. They weren’t sure, like, how it would, you know, benefit them, other than the sense that they were like, we know that we need to be better with organizing our documentation is sort of what they wanted to do. Well, right after we onboarded them, which, as Justin said, is a super quick process, they had a quick turnaround of support staff, and so all of a sudden they’re like, oldest member on staff was like, had been there for like, three weeks, and they said that they the timing of it was really serendipitous for them, because they didn’t miss a beat. And so they were able to use it, not only for onboarding for the new staff, because they had already started implementing putting every all their documentation in one place, and so they could just easily find what they needed to learn the information, but also that staff member could easily answer questions when people called in right away, and so so we knew that it had this capability of being able to be a very helpful tool with onboarding and training, but to hear it like actually happen for what might be considered an emergency, but for them to hear that they really benefited from that was, was great. So that’s one of my favorite stories that I have

 

Katie Smithson  12:07

heard

 

Shannon Baird  12:07

in the field.

 

Katie Smithson  12:08

Yeah, that’s very cool. And then can either one of you talk a little bit about So, I mean, I think correct me if I’m wrong, initially you did develop Navigator more as like an internal staff tool, but that has also evolved, and you do have member facing component as well. So do you want to talk a little bit about that? And has there been any, I guess, hesitation around MLS is adopting that it’s, you know, AI is still kind of, I don’t know it’s, it’s a little bit of a, I don’t want to say scary topic, but it’s, you know, I think a lot of people just don’t know what they don’t know. And so anything that we can do to, like, provide a little bit of comfort and ease that you’re doing the right thing by using AI break time out if you’re ready to put AI to work for your staff and your membership. It’s time to check out Navigator. REcore is the exclusive reseller of Navigator by Lindy head to REcore.net, forward slash Navigator, to learn more. All right, back to the show.

 

Justin Lundy  13:17

Our plan was always to allow MLSs and associations to use this however they want, including member facing though we have a process, we need the staff to utilize it first, because there’s no way that your documentation is up to date or correct or covers everything your members might want to know. And so as best practice, even if an MLS wants to make Navigator member facing answer the phones be the chat support or even text, you know, back and forth with their members, first the staff needs to onboard. They need to upload everything that they’ve got, all the documentations, rules, policies, procedures, anything that might answer the questions that their members ask, right? Because the last thing you want is to call Navigator and ask a question and it’s like, I don’t, I don’t

 

Katie Smithson  14:15

have an answer. And it

 

Justin Lundy  14:16

won’t lie to you that’s, you know,

 

Katie Smithson  14:18

yes, I was just gonna say, I think that’s an important fact to point out there too. This is why it’s so important to get the documentation right. Is because Navigator does not hallucinate answers. It is going purely on the documentation provided by the MLS or the association. And so when members call in for a question, a support issue, or whatever it’s looking at the documentation that is provided in Navigator. It’s not just like searching broadly and saying, you know, I don’t know. I can’t think. I can’t think of it. Yeah, it’s not. The

 

Justin Lundy  14:52

problem with llms is they answer for the sake of answering. You know, they’re built to answer. And so the way that we’ve kind of locked. That down is, if it’s not in the documents, it’ll say it doesn’t know, right? And so before you roll it out to your members and they experience, I don’t know, I don’t know, I don’t know, you know, for whatever they ask, give it to the staff, and you’ve probably got a lot of staff that know all the answers already, right? But it’s important that no matter what question comes in, whether it’s a phone call or a support ticket, you run it, they run it through the Navigator first all of them, even if they know the answer, because then, as the Navigator says, I don’t know, you can see those holes in your documentation. You can go into the analytics and sort by the I don’t knows, right? And fill all those holes and do that, do that for a few weeks, and until you feel very comfortable about, you know, holy cow, this covers a lot of things. And then in the event, because you’ll never get it all, in the event that they do ask some question that’s never been covered, the Navigator will simply hand them off to support. If it’s during operational hours, they can do a live handoff that that support person gets the whole context of the conversation with that handoff, so they can just pick it up and keep going and and this is where you know support staff can never be replaced. Is there’s always these fringe kind of edge cases that you’re never going to train AI on, because they come in and they’re so wacky or random that you’re just like, oh, I, I can handle that right as a human, but, yeah, good luck with AI. And you know, if it’s after hours, it’ll create a sport ticket for them and let them know that you know your your team will get back to them. So we you know, it is available for member facing,

 

Katie Smithson  16:43

yeah, there’s

 

Justin Lundy  16:44

a great resource for that. It’s going to save your team a lot of time, especially on those just, you know, same questions over and over and over

 

Katie Smithson  16:53

again, yeah,

 

Justin Lundy  16:53

yeah. And so I do think it’s great for that. Just, we don’t recommend running straight to,

 

Katie Smithson  17:01

yeah, just going straight into, don’t just dive in the deep end. All right, I want to shift gears a little bit talk about the relationship between REcore and Lundy. I obviously know why I wanted to work with Lundy, but I just want to, I want to talk to you Justin about, why did it make sense to you? What was the inspiration? I mean, besides getting to work with me on a regular basis, yeah, why did you want to work with

 

Justin Lundy  17:35

well, you know, first of all, we’re kind of joking, but like you have the deep network, right, right? And so does REcore. And what you’re building for the industry is something that that was already, you know, missing, and so you’re already providing a lot of value. You also have a couple products that Navigator ties perfectly into, right? There’s the assist product that you have, and your comply product, which you know you know more about than I do, but the time and Navigator is perfect, and it frees us up to just focus on the product while you help scale it across the industry. And I find that that’s, you know, it really is a perfect partnership for

 

Katie Smithson  18:15

that,

 

Katie Smithson  18:17

yeah. I mean, and I was just, I was very excited because, I mean, like you said, so we already core, we have two, two products that I think will pair perfectly with Navigator, which is assist, does support ticketing, and then comply, which is obviously compliance work. And so those are two departments that would benefit from using the Navigator. And again, like you said, you can, you know, if you’re using it for member facing support, it can tie in and create a ticket. And so that’s one of the things that we’ve been working on between our two teams to make that integration very seamless. But I just, I thought it was very exciting, rather than, you know, REcore trying to go out and build a similar thing that would compete. To me, it just makes sense to have this partnership where we’re both benefiting. And, you know, I think it just it makes for a strong relationship. And like you said, we we both have our different networks, and I have been in this industry for a long time, so a lot of people do know me and most the time take my phone calls, so that’s always a good thing.

 

Justin Lundy  19:25

Yeah, I mean, at the end of the day, both of our goals is to get the industry solutions that help it operate better and more efficiently. And like you said, instead of you trying to build something similar, or me trying to scale a sales team, both of those things take time, right? And they take trial and error, and this is already, like, you’ve got your established benefits, and I’ve got my products that you know work well, and have been tested, and so it’s perfect. It does help us to kind of scale quickly,

 

Katie Smithson  19:56

yeah, yeah. It’s been, it’s been really awesome. I’m working with y’all and I mean, I think one of it’s just, it’s always impressive to me, how quickly you take feedback from customers and then turn it around and make it into reality. That’s always super impressive. And I just, I love the Navigator. I use it sometimes for contracts. I’ll upload it, upload a contract in there and ask questions. So that’s my little secret that I’m sharing with everybody.

 

Justin Lundy  20:25

Yeah, it’s great for

 

Katie Smithson  20:26

that.

 

Justin Lundy  20:26

It’s funny, like we so some of you may know we built the best practices Navigator for cmls.

 

Katie Smithson  20:36

Yes,

 

Justin Lundy  20:37

I apologize for this story, but when we rolled it out, their their team had, you know, made sure they’d gone through all the documentation that was going to be behind and powering the best practices Navigator. And it works just like the regular Navigator, but it’s, you know, got best practices for all sea and losses members. And the first, probably the first person to try it was Mike warser, and he asked it, what, what is the MLS good for? And compensation?

 

Katie Smithson  21:10

Oh,

 

Justin Lundy  21:11

like, and it was like, Oh no, it’s, it’s broken. What did we do? And so we’re, we have this fire drill on our side, like, what is Why is it like, what is it doing? Why is it answering it like that? And it and it’s like, well, let me ask the same question, and then it answers. It’s like, you know, all of the value, including compensation, but it has a citation link, so we click on the link, and it takes us to this old white paper that’s living in their documents.

 

Katie Smithson  21:42

It

 

Justin Lundy  21:42

was like, sunny, hey, this is why that’s happening. And she’s like, Oh, my God, we come through this database so many times to get rid of all of this old documentation, and we still missed

 

Katie Smithson  21:57

one.

 

Justin Lundy  21:58

And it’s like, well, if you’d use the Navigator,

 

Katie Smithson  22:00

you’d actually used it before you turn it out.

 

Justin Lundy  22:04

It was, that was a fun little fire drill. But, like, yeah, if like, it’s, it works on, it works for a number of things, just just figuring out, like maybe even what to get rid of,

 

Shannon Baird  22:25

yeah,

 

Katie Smithson  22:26

yeah.

 

Katie Smithson  22:28

All right. So to wrap up here, just what, what does the future of AI in the MLS world look like? And I mean, if you want to speak specifically to some things that you are working on. I think that would be awesome. Hint, hint.

 

Shannon Baird  22:48

I think that

 

Katie Smithson  22:49

one whoever wants to talk about,

 

Justin Lundy  22:54

it’s funny. I for a long time have believed in voice technology, if you couldn’t tell and what, what’s happening in front of us is, you know, everybody’s jumping in on this AI and voice kind of bandwagon. But I think where a lot of them are misunderstanding what voice means is they think that talking to your technology, putting a microphone on it while it transcribes what you’re saying into like a sentence, and then sends that into the product. That’s not what it is. What’s happening right now is more along the lines of agentic assistance are being released into the market. We’ve thought for a long time, and we’ve been working on the ability to not, not like, transcribe a voice command and send it into the tech. It’s more that the technology itself understands what you’re saying and can then go do what you told it to. And with that in mind, I mean, the MLS is will never get there. The vendors, the MLS systems providers, they’ll never get to a point where you can talk to their technology in a way that’s meaningful. But fortunately, with their already established not just the ML systems providers, but all the vendors you know in this space, whether it’s MLS products or agent products or broker products, they have already done the work to build reliable platforms, and if they just expose those endpoints to the right companies, agentic AI can tie into that perfectly and give the agents and the MLS staff something that we’ve only seen In like Star Trek, right? Or in Iron Man, where you can literally talk out loud. You’re like, have a really meaningful conversation with your technology. And it’s not just like, telling you, oh, that’s a good idea and blah, blah. Like, not that kind of conversation, but like, Hey, I’d like this to go get done. And. Like you got it, and then off it goes, right? That’s what’s actually possible now, and that’s what’s happening. And so I think the mls of the future has to kind of understand that and harness it, and understand that their agents and members, that’s what they’re going to expect. Like, we’ve seen this kind of viral, you know this, this viral uprising of open cloth, right? And others, others like it, where you kind of just plug this AI into your entire life and let it go. You just kind of tell it, hey, this is how I normally operate. You’re going to do that now and then. It does it. It’s very unsecure at you know, it’s it’s not, I do not recommend anybody

 

Katie Smithson  25:47

just, but

 

Justin Lundy  25:54

it’s happening,

 

Katie Smithson  25:55

yeah,

 

Justin Lundy  25:55

like at an alarming rate. These people are just get, handing it the keys to everything, because the demand is so pent up, and everybody’s like, we’re so close to this future that we’ve always seen in like sci fi movies.

 

Katie Smithson  26:10

Yeah,

 

Justin Lundy  26:10

it’s it’s literally right in front of us, and people are just making that extra leap unsecurely to get there. And so that is why we’re building Nora, that Nora solves that whole problem. She ties into every product that is either in your MLS or an agents, you know, tool belt, that they have already paid for. And they’re, they’re having to manually log into each day individually and share data between all of them. And, you know, just the mess that it is Nora now ties into all of them, and they simply talk to Nora and she goes and executes on their behalf in a secure way. You know, security is our foundation. So,

 

Katie Smithson  26:51

yes, that’s

 

Justin Lundy  26:52

where I think it’s going. And I think this, as slow as this industry is to adopt things, it won’t have a choice,

 

Katie Smithson  27:01

yeah,

 

Justin Lundy  27:02

either they’re going to be proactive about it and offer their members solutions that are safe, or their members are going to grab on to things that are not safe,

 

Katie Smithson  27:10

right?

 

Justin Lundy  27:11

And either way, they’re going to have

 

Katie Smithson  27:13

them. Yeah, and I’m, do you think that, like, I mean, in my head, that just poses a threat to the MLS and the security of that data and the things that an MLS does operationally. So I would, I would imagine that anybody listening to this would want to have products that have been vetted and tested and secure, as opposed to just opening it up to say, like, okay, yeah, you can, you know, go plug in whatever AI tool you have created or accessed.

 

Justin Lundy  27:45

Yeah, yeah. I mean, hopefully they’re already thinking about that. Hopefully their staff has been informed not to plug in their life, because it may be too late for some of them. But yeah, I think it’ll be become more clear. I may be talking like, it may seem like I’m talking way down the road, but it’s gonna be

 

Katie Smithson  28:08

this year. Yeah, yeah.

 

Katie Smithson  28:10

I mean, I like technology is moving so fast these days, and I don’t, yeah, I don’t think it’s a matter of, like, years that we’re talking about. It’s much shorter, like, months. Yeah,

 

Justin Lundy  28:24

a lot of this is already built,

 

Katie Smithson  28:25

yeah, well, I am excited to see what y’all do next. I am very excited about Nora and how that can just further facilitate the use of the Navigator. I mean, I, I I just want to emphasize that, like you said, Justin, I think both of us are looking to solve problems for MLSs and associations, down to brokers and agents as well, just to make everybody’s lives easier and using AI in a way that is practical and helpful, as opposed to just, you know, hey, chat, GPT tell me what to wear today. Yeah,

 

Justin Lundy  29:07

yeah, what should I do? I don’t want to I don’t want something to tell me what I should do. Right? Go do it.

 

Katie Smithson  29:12

You want to tell it to do what you need it to do. Yeah,

 

Justin Lundy  29:15

that’s right,

 

Katie Smithson  29:15

very cool. Well, thank you again, both for being here today. This was super fun. We’ve had a lot of good laughs. Probably a lot of it will be edited out, but that’s okay.

 

Katie Smithson  29:29

Thank you so much.

 

Katie Smithson  29:31

That wraps up this episode of REpod. If this conversation made you think differently, don’t keep it to yourself. Subscribe, share it with your network, and sign up for our newsletter to stay in the conversation. The future of organized real estate gets better when leaders engage.