Build or Buy? Legal AI Needs More Than Fast Code (Full Transcript)

[00:00:00] Speaker 1: Hi, I'm Zach.

[00:00:03] Speaker 2: And I'm Stephanie. Welcome to the Lawyer's Podcast, one of the many ways we help lawyers build healthier law firms, better businesses, and more sustainable lives. Today, Zach's talking with Nick Coleman about when to build and when to buy in the time of AI.

[00:00:20] Speaker 1: It's an interesting conversation. Nick has some, he has built a platform for himself and some other attorneys and you know, in the, in the land of SaaS is dead and we can spin things up pretty quickly. When do you, you know, when do you build your own thing? When you, when you get in somebody else's stuff.

[00:00:36] Speaker 2: So yeah. I'll be curious to hear, I know what I just told my husband recently when he brought his big idea, like, Hey, I can code this thing. And I was like, no, you can not, you do not know how to code and why you can do it with AI. It doesn't mean you should. So yeah, I think that's because he doesn't understand security and data and like, I mean, Come on guys. There's some complicated stuff there that I'm, I'm like, let's not play around.

[00:01:00] Speaker 1: You are, you are speaking to what we talk about. That's exactly right. Exactly right. Stephanie. I, this last weekend I ran a 5k with my cross country team. One of my, one of my athletes puts on a puts on a charity 5k every year and she's been doing it since she was a freshman. She is the one that organized this. She's a senior now. It's, it's absolutely amazing what she's able to do. The team does the 5k and I went and ran with them and it made me think because I'd literally gotten to this situation. It made me think of a rule that I have with these, with these athletes and I tell them, don't change speed going uphill. I always tell them you can change speeds. You can, you can walk, you can run, you can do things how you want to, but don't change speeds going uphill. And what I mean by that is bottom of the hill is you feel good. You feel good about yourself. You know who you are. You're, you're, you're kind of a better person at the bottom of the hill in the middle of the hill. When things get difficult, you're making a decision based on the person who is under stress. They, they are, they are having, they're going through the middle of this difficulty. Why not trust the you that is pre stress or non stress you. So go the speed that you decided to go at the bottom of the hill. When you get to the top of the hill, whatever you want, change speeds, walk, whatever you want, but don't change speed in the middle of that hill. And obviously that has normal human implications as well.

[00:02:34] Speaker 2: Yeah. I need to think about that. I'm sitting here thinking, but Zach, I might not make it up the hill. Like I feel like I'd be the one that's like, no, I actually need to walk.

[00:02:44] Speaker 1: This isn't, sometimes I think that is, that is a real critical thing here. Trust the person at the beginning of the hill, um, bit off more than they could chew. Sure. Sure. If there is a real, real reason to do it, but I mean at some point it's like, then fail, you know, like, like don't change, don't try to change speeds. You might fall, you might, you know, but don't, don't get in the habit, oh, things got difficult. Let me reassess. Don't reassess just because they got difficult. Yeah.

[00:03:15] Speaker 2: That totally resonates. And I do like that. And where it also made me go as leaders, we talk about this in our book and next level leader, you know, we show up under stress differently for our team and we don't always realize it, but like, how are we showing up if we're under, as things are difficult, we're under a lot of demands and stress to our good qualities can sometimes get exacerbated into really bad ones. So like if I'm good under stress and I'm like, I'm decisive, I can make decisions under stress that could look like being demanding and critical.

[00:03:48] Speaker 1: Yeah. Yeah. I, and I think it's like underlying that whole kind of uphill thing. I think it's important to understand the different use that are, that are out there, the different forms that you can take based on what's going on, you know, external factors, what's going on around you and learning to cope with those things, learning to, to, you know, wrangle those different Zach's that show up to the party.

[00:04:15] Speaker 2: I love it. Let's just in there. That's a good nugget of wisdom that you're dropping on everyone today. So now let's check out your conversation with Nick.

[00:04:35] Speaker 3: My name is Nick Coleman. I am the CEO and founder of LexMed also recovering social security disability attorney. And what we do is build tools to make social security disability attorneys, better advocates. And essentially LexMed it's been a passion project built out of the frustrations of me as a young attorney on the road, handling hundreds of cases a year while making peanuts. And so, yeah.

[00:05:04] Speaker 4: So Nick, you and I've talked a couple of times about LexMed and about your journey to building LexMed, but this is, this is a SAS platform that, you know, you built for yourself initially or to help yourself move forward using artificial intelligence. So you looked around and you said, Hey, here's some tools that can make me not only faster but more effective. Talk to me about, about the, the impetus of building LexMed.

[00:05:35] Speaker 3: Yeah. So I've always been a tech nerd. I don't really technically have like a technical background. I'm not, I'm not going to write code, Python, you know, understand all the syntax, but I'm a systems builder. I have understanding of leveraging technology the best way I can. And I'd always had build macros and use different plugins with Adobe PDF to help kind of streamline my workflows. Right. Yeah. Before AI was AI, before 2022, OpenAI came out and released Chats B2. The things that I wanted to do, I really couldn't do other than using the tools that were available to us. And with social security disability, a lot of everything I do is really dig into thousands of pages of medical evidence and figure out how those medical conditions link to relevant regulations and ultimately how do they explain how this person can no longer work, which is just a very time consuming task. So a lot of what we were doing is just doing, we'd say control F on steroids and matching words. Right.

[00:06:38] Speaker 4: And then I would build scripts to find those matching words and flag them. Right. Yeah. I guess one of the things that we've talked about, like in that we're talking about a product that is kind of enhancing what somebody does, speeding them up, doing something that they do more quickly, but you and I have discussed, and I love this aspect of AI, the ability of this product that you've built to do things that you wouldn't even do if you didn't have this sort of speed on your side. Talk to me about that. Talk to me about some of the things where you're not only saying, oh, we went faster, we went more quickly, it eased things up, but it's, I did better because I could do something that I wouldn't do.

[00:07:22] Speaker 3: What our systems do is allow you not only to comb through a record, we're not just this AI, we summarize it for you, right? We're actually doing detailed analysis while we are breaking out the most important aspects of someone's longitudinal medical history and actually grasping and picking up patterns that are very easily missed when you're running a volume practice on tight margins, right? And so you might have 30 hearings in the aggregate where you have over 300,000 pages of records and you're expected to know every needle in the haystack, right, with each case. And there's only so much time in the day, right? And so even the best attorney is going to miss things because there's just too much information. What our system does is digest that information, draws the connections into creating patterns of what's happening over a longitudinal history that might be 15 years of medical treatment, but also looks at how these things connect to the regulatory framework, which the social security uses to evaluate these cases, which is often built on ICD 10 codes. You think of like when you go to the doctor, how they bill your insurance company and you're supposed to have to know, you have to play kind of junior doctor and an attorney, right? When you're evaluating this stuff and it's very time consuming and it's very technical and to be good at it, it's difficult with the amount of volume that you have to do.

[00:09:06] Speaker 4: So talk to me about some of the things, some of the specific things that we've discussed on. When I look at a lot of AI, it speeds me up. Great. Okay. And then we have a question of, is your practice an hourly billing practice or is your practice a contingency fee practice? And speeding me up in some cases doesn't help me. And yes, so obviously speeding you up helps you in a contingency practice like social security. But you've also said that it does, you know, you've been able to find things that wouldn't have even been done, you know, like, and get things done not only faster, but better. And yes, more quickly, but like, because we, we almost cut the cases legs out from under

[00:09:50] Speaker 3: it. A hundred percent. And see, there's one drastic thing in the way of social security disability practitioner runs a practice, a contingency fee to personal injury. The numbers just aren't there to spend the amount of time that you would spend on a case that's a trucking case is worth $3 million. You spend hundreds of hours on that because at the end of the day, if you win, it's big social security. There's some nuance to the fees, but really for the most part, people were getting capped at $9,200 a case, right? And so you just, you had the economics of it. Don't make sense for you to spend hundreds of hours on a case. So I'll give you the prime example. And what happens is, is when these cases, after they go before an ALJ and you lose, there's a tribunal that goes, there's above them that you, so I'm going to interrupt you.

[00:10:39] Speaker 4: I love, I love that concept because this is like attorneys get this of like, you're going to go in front of the ALJ. You're going to lose. Fine. Like that's just the step we have to go through. So yeah. Sorry to interrupt.

[00:10:50] Speaker 3: Yeah. And then you're, you're in this, what I call the administrative universe. Before you can get out and sue the administration, you've got these four stages. You go before the judge, that's the third stage. You lose. And then there's this tribunal that you'd notice nobody sees that, that you appealed to in Virginia, the appeals council. They released their numbers every year. And on average, in the last 10 years, there is a 15% remand rate, which is horrible. It's this, you have no chance. You can go to federal court, it goes up. So here's the scenario. Jane's case has just been denied by an ALJ. It's hits your, it's hits your desk. You're on day 65. That's the deadline that you've got to get this in. You also have five cases that are gimme's on Friday, but you still need to prepare for them. Right? Well, Jane's case is a difficult, uh, what they think is fibromyalgia, but is also really possibly a more complicated degenerative muscles condition. You've got some evidence and it's probably gonna take some research and really to draft a well-drafted brief. It's going to take you eight hours. Well, are you going to spend eight hours on Jane's case or are you going to dedicate that time to those four cases on Friday where you're, if you just put a little bit of time, you'll win all of them. That's the economics of it. And so Jane's case doesn't get, doesn't get taken care of and call it in and it's a garbage

[00:12:16] Speaker 4: brief that's submitted. To be clear, Jane doesn't lose. We just extend the time considerably that we're going to be doing Jane's case. Yeah. Yes. Like considerably. But if you had that moment, if you had that, that capability, then you can wrap that one up too. I think that's, that's the thing that's interesting to me is the, the economics of that, the almost the hours economics of that of like, okay, well, if I can write a really sharp brief on this, which would take me so much time, I can get this thing done right now. And Jane can get, get this, you know, handled more quickly. But like we, we, we have to, we have to do the other ones too, where, you know, we're doing the right thing. We're making the right choices. But now with, with some technology, I mean, we can adjust our choices, you know?

[00:13:07] Speaker 3: And what it does is it honestly levels, uh, it's kind of, kind of become cliche and cheesy like democratizing justice. And what, what we're really trying to do is make you, the attorneys be able to effectively advocate equally amongst all of their clients, right? And that's not, that's not what's happening because Jane's case, yeah, you know, she's got some serious conditions, but there's, it's, there's a lot of nuance to it. It's going to be difficult, but the economics of it don't work for you to spend as much time on her case when you've got four winners that are going to produce maybe over $30,000 in fees in a month, right? You're going to give some time to Jane, but not the amount of time that it deserves.

[00:13:53] Speaker 4: That's a fascinating concept when you're, when you're talking about, uh, democratizing, it's not democratizing between, um, you know, one side and the other side, it's democratizing it across your clients. Yes. That's interesting. Okay. So the, I definitely wanted to talk to you about the building of this, the reason you build it, you know, and there is the aspect of, we want to go faster. We want to do things better, but you built this yourself for you. And um, when you looked at this and you said, Hey, I've got this great product. You also said, well, I want to turn this outward towards other people, sell it, let other people use it. I think that's a reasonably familiar idea for attorneys. I get, honestly, I get people calling me all the time saying, Hey, this is a, this is an interesting use case. I'm, I'm an attorney who's done blank. And so talk to me about, about making that switch and turning the, the, uh, the product from internal to external.

[00:14:54] Speaker 3: I always, I've kind of had a knack for this in this practice area. I started out in just volume doing administrative work and then, uh, migrated to doing the appellate work, which is more technical. So actually doing litigations and courts. So I ended up becoming the go-to guy in Arkansas to send referrals for the appellate work. Cause a lot of people were intimidated by going to actually doing, you go from administrative work to actually getting federal courts. A lot of people are intimidated, intimidated by federal, I mean, it's not just social security.

[00:15:23] Speaker 4: It's any, it's any, it's not where they're comfortable. It's not where they've been. It's not where their wheelhouse necessarily. And there's nothing wrong with that. It's just not their wheelhouse.

[00:15:30] Speaker 3: Yeah. So I kind of became a name myself just because of, I was younger. And then I was, I was handling, uh, one of the few just solely appellate practices in Arkansas and I would just have so many different conversations with attorneys like this is, I would love if I had this, I'd love to have that. And there were things that I was even doing, like I said, before open AI came out and there were things that I was actually working out where we're on the cusp of open AI or using AWS medical comprehend some of their natural language processing stuff before. And I was just hearing all this feedback from my people of what they wanted. And then when open AI came out, it made things that I was trying to do using regular expressions, natural language process and things like that.

[00:16:13] Speaker 4: I got rejects.

[00:16:14] Speaker 3: Just blew the doors off of it, right? Because what we're trying to do is connect the dots between very technical medical data and over complex regulatory requirements, archaic things. Right. And so the big thing was like, I'll give it a short example and I've get too technical, but like what you're trying to find is one regulation to establish disability requires them to meet a musculoskeletal listing for the back. But the word is limited range of motion. With regular expresses that are, you know, NLP, it really had to be a one for one to match that from the medical record. And I changed that because it can distinguish between, you know, like words between the same thing and still get you a conclusion on whether these particular conditions meet these requirements. And so that's so real.

[00:17:04] Speaker 4: For example, we're saying instead of it just looking for literally the word limited, it's maybe reduced.

[00:17:10] Speaker 3: Yes.

[00:17:11] Speaker 4: You know, it puts that into the bucket of like what we're going to flag.

[00:17:15] Speaker 3: Yeah, that would a lot because in it, every electronic medical record has a template of how they use particular language to describe symptoms or objective findings within a record. And with the way it was very difficult to do it before. Now with AI knows it can distinguish between different words to make that match. And so that's what when I was like, oh, this is big. And then so I started building just when it was like three point, was it chad GPT three point five?

[00:17:46] Speaker 4: Right. Oh man.

[00:17:48] Speaker 3: Okay. Yeah. Yeah. It was like, yeah, I guess that was like five years ago. So I was like, I started creating just prompt engineering, all of this stuff. Right. Like at the time, I was like, this is amazing. And then I started doing some like thought leadership talking in this space. And then people were like, man, this is really. And I kind of had the idea that I wanted to do this, but people were like, I want this Nick, will you build this? And then I started kind of tinkering around some more. And then I just started building tools across the whole life cycle of social security. And then I started having products and, and then, and then it became, and then I've tried to get a company off the ground, but I was actually running a practice and moonlighting as a tech founder at night and it didn't work. And this is where it comes to like, dude, why don't I just vibe code this stuff myself?

[00:18:36] Speaker 4: I think that's, I think that's the next question is, you know, you, you've, I don't, I do want to call this vibe coding, but I also don't want to call it vibe coding because you know, there's a, there's a big, there's a big spectrum of what I consider vibe coding. You know, some of it is like, Oh, I just threw up a, an HTML you know, visualization over here and we call it vibe coding. And then some of it is, you know people designing something, vibe coding it and then sending it off to engineers to like really, really sharpen it up. So this really let you design all of this stuff in a, in like test it and kick it around and all that. And I think the question then is, yes. Why would somebody, and you have to know, you have to think about this as a founder, why would somebody come to your product instead of just building it themselves? Like for me, what, what, when do I decide to buy versus build?

[00:19:31] Speaker 3: I think there are several applications of using AI where you don't need to have the whole SaaS platform and it's small one-off cases. But with what we're providing, there's so many layers to it, whether it's security to regulatory maintenance to something we'll talk about later, as far as like getting feedback and getting community into, into making these product, product, products progressively better. It is just so difficult to build out a product like we have in our platform because Murphy's law is alive and well. Things go wrong every day, every day they go wrong. Things that we don't know until they go wrong, until a customer tells us that they go wrong.

[00:20:18] Speaker 4: And if you're in court, when a customer tells you that something went wrong and they need it, you know, recoded, what are you going to do is just Zach Glaser sitting there in court vibe coding up a SaaS platform. Maybe it's not a customer, maybe it's your assistant.

[00:20:33] Speaker 3: All of it seems because these, the, those frontier models have become so good at coding and, and, and, and prototyping essentially, it seems like you can just run with these things and do everything. The thing is, is like, yes, it might create a good prototype, but are you going to have the time in the day to keep up with the maintenance of keeping all the libraries up to date on that program? Are you going to be able to keep up with that regulatory updates? So security updates their programs all the time. We have a system where we have two SMEs right now, including myself, where we stay on top of this to make sure that all of our regulatory requirements are aligned with what's present. And I just don't, I've, I tried to do this, build this company as a solo practitioner and tried to moonlight it and just did not work. And I just can't imagine another busy practitioner being able to do, build software that they can rely upon and trust for, you know what I mean? It seems, I do.

[00:21:44] Speaker 4: So I used to tell people when, when, you know, when I would meet with attorneys and they'd be like, I am, you know, building this internal product and it's a, it's an automation or I'm, I'm the one that sets up my computers. I was the guy that set up my computers and the infrastructure in my office. Great. And I used to tell people, you know, you can change your own oil in your car. Like I, I've got a big two 50, I can change the oil in my truck in a heartbeat. Like I don't even, I don't have to lift it up. I just slide right under there, but I don't do it because why? Like somebody else can do it cheaper and better. The only reason to change your oil in your car is if you like doing it and then it becomes, then it's a hobby. Then it's not, I'm actually working on my business. So if you're setting up the computers and the infrastructure in your office, great, do it. But that's a hobby. That's not, that's not you, you know, running your business. And I, I think now we're talking about that's a full functioning business, separate business. This is what you came across. It's like now you're not just changing the oil in your car, you're setting up a quick lube in your garage and you're trying to change other people's oil. And at that point it is, it is a fully functioning business. So how do you do that? Like, and you found I can't run this business and practice well, conversely, how do you practice and run that business? And I think, I think that's difficult. But what I liked about what you started with was if you're just creating tools, small one off tools in your office, fantastic build that great, you know, like, like ease into that infrastructure. But if you, if you need it to be a big functioning machine, I mean design it, but maybe ship that out or buy it, you know?

[00:23:33] Speaker 3: Yeah. I mean, I think there's probably more opportunities for solos and things to like do that. But if you're wanting to put this to where you need to have a program that needs to follow internal governance with AI, have other employees use it. I think that you're going to have a lot of headaches unless, unless you can hire an on stack, like internal developers, which gets very expensive, can handle the cloud expenses involved with it. All of that. Then you think you're going to, I think you're going to run into a lot of headaches and you're going to end up spending a lot of money trying to build something where you thought it might be $50,000 and it turns into a million dollar project and it never goes anywhere. And it happens.

[00:24:18] Speaker 4: Yeah. Yeah. I, I absolutely. And I think it's happened like that happens even prior to AI, just think that this question is really here. It's been souped up because of the ability to spin things up so quickly. Like we can, I was talking to some people at a conference last night some, some CEOs of some of these legal tech companies. And it's, if you can, if you can build at the speed of thought, you know, like we can build at the speed of thought now, like, how do we, how do we maintain this stuff? How do we, how do we like, we're going to get behind the eight ball if you just don't have experience with it and don't have the ability to kind of focus, you know?

[00:24:58] Speaker 3: Yeah. And there's just so many nuanced things that I didn't even appreciate with product development that like, that I do appreciate now because I'm in the mix of it. Right. It's like, you don't know what you don't know.

[00:25:08] Speaker 4: Yeah.

[00:25:09] Speaker 3: Right. It's like, you don't, with all these security measures, like it just, just privacy, like AI has been great because it, like we've said, it makes things so easy, but it's an illusion. It's a loop. It's illusion when you actually start trying to implement this in something that you're going to use day to day.

[00:25:29] Speaker 4: Right. Yeah. So one thing I kind of wanted to get to as well is the, when you can spin up an S and a SAS product, like if, if let's say I don't want to sell it to other people and I do want to spin it up internally and I'm not super concerned, you know, I'm, I'm able to deal with the, the security and I hired somebody in house. I've been thinking about the idea of like, what does a SAS product bring to the table? A collective SAS product, a public SAS product. What does that bring to the table that a local product that I built myself doesn't and it's community. A hundred percent. It's, it's, you know, it's you and your other SMEs in the background, you know, making, keeping things up to date. But even more than that, it's the, the user feedback loop that you can create in that scenario. Talk to me about that. I know you've had experience with that user feedback loop.

[00:26:23] Speaker 3: Yeah. Yeah. It's, it's been great for us. You know, a big thing with us, you know, right out of the gate is obviously it's, it's, it's privacy and security with all attorneys, but we're dealing with so much such sensitive medical data all the time. It's always at the top of mind, right on, on how we are using claimant information. The big thing right out of the gates, obviously like ZDR policies and, and ensuring that we're following best practices so you, you, you can be in compliance with model rules of professional conduct 1.6. All that stuff. Right. And so we, and, and on training data and training, using, using claimant's data to train. And that's like going to first questions. We are not training our models like in a traditional sense, we're using customer feedback to make our systems better. Right. And so as we've explained that to our customers they understand that it's not just their feedback is more critical than maybe what they've provided feedback to their case management software program. Whether it's like my case or smoke bar or whatever, you know what I mean? Like I've used it, you give me, it is more critical because although I like to think I know everything about social security disability, I don't. And there are going to be edge cases. There's a million edge cases with the law. Right. And so, and it's things that AI cannot, that's one thing where AI is not going to be able to know everything. It can suck up the whole internet, but it's not going to know every age case, whether it's social security disability, you're dealing with a personal injury or mergers and acquisitions, whatever. There's too many edge cases because it's, it's human interaction. Humans are unpredictable.

[00:28:08] Speaker 4: Right. And, and AI is not creative. You know, it may seem, it may feign creativity, but it's not creative. So it's not going to come up with a novel solution to the problem like a human will.

[00:28:20] Speaker 3: So a big thing that we try to establish early on when we onboard with our customers, and I think it helps that I'm an actual practitioner too, when I'm telling these people, because I told them, you know, I hate all the downfalls of these SAS companies I've dealt with in the past. And so you really try to build a connection. It's like, we're not going to do that to you. When we ask for feedback, we mean it because what you give us is going to fundamentally change how we build this product and the things that you see, we're going to fix that week. If there's a problem with it, not detecting this issue that is critical for your case, we're going to fix it right then. And so we have this feedback loop with all of our customers where they, they feel heard. I won't name the companies, but I would tell these companies things for myself and it would just go into a void. Great. Just attorneys complaining again about our product. I don't see it as complaints. I see it as the, you are fundamentally making our product better for us. And then for, you're not just making the product better for you. You're making that product better for 20 other attorneys across the country. And when you make the product better for them, it doesn't just have the impact on the attorneys as a downstream effect, because everything we're building is impacts a claimant that is applied for social security disability benefits or has been denied benefits. And our team buys into that too. And so that's the community we make with our attorney customers. And I say this, I was like, we don't, this, some of these things I say, you're seem kind of cheesy. Look, we're not that we do not look at this as a customer vendor relationship. We really look at this as a partnership because we want to make these products work for you. We want these products to work for your clients. And so it's vital for us.

[00:30:17] Speaker 4: I think you've hit on something there as a founder that the SAS products writ large are going to have to make sure that they address from my perspective is that if I can build something, if I can build a product that doesn't listen to me, I can build a product that doesn't listen to me all day. I can build a product that doesn't update and make my life better all day. But I think there's a heightened importance on that user facing user feedback, that loop that we have to be building this to meet the user where they are, but also meet the user, not where we want them to go, but where they want to go.

[00:31:00] Speaker 3: Yeah. And then I was about to touch on this. There's a little bit of like having some humility to about building these tools, right? Because you went like, I was kind of joking. I think I know everything. I don't know anything. I know I did really very, very well for myself in the social security disability practice and I'm pretty tech savvy for an attorney. So you do kind of have an ego about yourself, right? And so the biggest thing for me was learning to listen, and once I've really kind of honed in on this and like, listen to these customers and let them kind of guide the product is really been beneficial for us and not get to, you know, get your feelings hurt because they didn't like it this way or whatever. Like that's easy to do, right? Because this is my baby. It's just like, it's just like with anybody who's building a product, right? They don't know what they're talking about, but no, you should read my YouTube comments. Yeah. When you get to, and I think that's with any founder who's, this is their, you know, passion project. Like there's a lot of emotions thought with it, but once you can kind of, there's, it's good to be proud, have a lot of pride about your products, but also it's also good to be able to look at this from a neutral perspective and say, this feedback is crucial. You need to take this and accept it because ultimately it's because we've done that and we've made ourselves have products that are really kicking butt in this particular market.

[00:32:29] Speaker 4: Love it. Well, Nick, I appreciate you sharing your experience with, with me and, and I mean like sharing your, your concept of, of this build versus buy with SAS products and like what the, what the, the benefit is, because I think a lot of attorneys, a lot of people are looking at this and you know, I met with Matt Spiegel the other day from the law Maddox and he was like, SAS is dead, you know, because everybody can build so much and you know, whether or not that's true. I think there's a, there's an argument that the community aspect of, of this and, and the maintenance aspect of this, like it's, it's not dead, but it's a different animal

[00:33:07] Speaker 3: now. It definitely is. Um, but I still think there's value in some platforms where, where, where you were the human element, like the way we are using it in our product development is still vital. Um, yeah. Yeah.

[00:33:21] Speaker 4: So no, thank you very much for having me on the show. I've really enjoyed it. Yeah. I appreciate it. And, and, uh, where can people find you? I imagine they can find you on LinkedIn and they can obviously go to lexmed.ai to, to check out, you know, the product, but they can find you on LinkedIn and yeah, I'm very

[00:33:36] Speaker 3: active on LinkedIn and then I can go to our website, www.lexmed.ai.

[00:33:42] Speaker 4: Sounds good. Once again, Nick, thank you very much for being with me. I appreciate it.

[00:33:45] Speaker 3: Thank you.