Insurtech at Speed: Engineering the Future of Specialty Insurance | The Pair Program Ep88

Speaker A: Welcome to the PEAR program from Hatchpad, the podcast that gives you a front row seat to candid conversations with tech leaders from the startup world. I'm your host, Tim Winkler, the creator of Hatchpad.

Speaker B: And I'm your other host, Mike Gruen.

Speaker A: Join us each episode as we bring together two guests to dissect topics at the intersection of technology, startups, and career growth. Welcome back to the PEAR program. I'm your host, Tim Winkler, joined by my co host, Mike Gruen. Mike, how's it going?

Speaker B: What's going all right? How you doing?

Speaker A: I'm good, I'm good. Yeah. Like, we're getting into, uh, Thanksgiving mode here.

Speaker C: Yep.

Speaker A: And so it always kind of leads to some controversial food takes, I feel like. So I thought it'd be fun to kind of kick it off with a little holiday themed, um, f. Marry Kill with some staples, uh, when you think about Thanksgiving food. So f. Marry Kill. You know, stuffing, you know, mashed potatoes or cranberry sauce.

Speaker B: Uh, well, it's definitely kill cranberry sauce.

Speaker A: Yeah. That was such a layup for you.

Speaker B: Uh, and then it's probably, um, uh, what was it? Mashed potatoes or stuffing or stuffing. Yeah, I'll marry mashed potatoes. Uh, yeah, could. Could live off that. That. That'd be my.

Speaker A: Yeah, we had. We have the exact same answers. I feel like that's a. Does that line up with you guys, too? It seems like a. Am I missing? Uh, missing.

Speaker C: Even before you said something, I wanted to kill cranberry, so didn't you need

Speaker B: to bring it up and need an option?

Speaker A: Yeah, some people are. Are pretty, you know, ridiculously obsessed with the cranberry sauce. Thanksgiving without it.

Speaker B: My wife and her sister seem to be obsessed with the cranberry sauce. Uh, so there you go. Um, this year, speaking of controversial takes for Thanksgiving, uh, I believe my mom put out an email suggesting that we were going to do Thanksgiving brunch this year. So, um, it's a little different. Yeah, it's a little different. We'll see how that goes.

Speaker A: Yeah. I mean, you know, my biggest pet peeve with Thanksgiving is, like, we always do the travel to somebody else's home. Usually it's my wife's family, and then I'm exhausted and have to drive another hour home. So, like, never really get the chance to just relax after eating.

Speaker B: Nice. Yeah, we, I go up, uh, we go up and then spend the night and then usually come back on, like, that Saturday.

Speaker A: Nice. Cool. All right, well, uh, today's episode has absolutely nothing to do with Thanksgiving foods, but we're, we are going to be dialed into an industry insurance, which I'd say a lot of folks are thankful for. And, uh, today's focus is going to be on the specialty insurance market and how modern technology is catching, uh, up to modern risk. And so to help, uh, us break some of this down, uh, we've got two guests. Nathan hall, the CTO at Ledgebrook. Fast uh, moving startup that's rebuilding how commercial insurance gets quoted, underwritten and bound using a modern tech stack, smart automation and a broker first mindset. Uh, and then alongside Nathan, we've got Mike Mansell, uh, a principal at American Family Ventures, one of LED Brook's, uh, earlier investors. Mike brings the VC lens on why infrastructure structure for startups or leading the next wave of insurance innovation. So always great to get these two different perspectives on the PEAR program. You know, Mike, Nathan, thanks for, for joining us on this one.

Speaker C: Yeah, thanks for having us.

Speaker D: Yeah, thanks for having us. Looking forward to this conversation and uh, talking about things besides, uh, Thanksgiving maybe.

Speaker A: I was gonna say they're talking about Thanksgiving foods.

Speaker D: I'm glad you didn't ask us that question around.

Speaker A: Well, uh, well, before we get into, you know, the main do kick things off with a fun segment we call Pair Me Up.

Speaker B: Uh, here's.

Speaker A: We'll go around the room, spitball a couple of things that go together or maybe don't go together. Mike, uh, you lead us off. What do you got for us? Uh, today.

Speaker B: Yes, so, yep. Uh, so my sister, uh, in law was visiting recently and just introduced me to um, edamame, uh, and wasabi, like various edamame covered with things. And the one that I gravitated towards, wasabi. Ah, covered edamame, which are, you know, soybeans. Um, and they've just become like a great snack food. Nice for protein and then the heat and the uh, the uh, flavor of the wasabi to sort of cut the.

Speaker C: The.

Speaker B: Just the monotony of eating soybeans. But it's been a good, uh, go to snack for some, Some quick, healthy protein.

Speaker A: Nice, dude. Yeah. Classic little food food pairing to kick it off. Yes, I. My daughter's obsessed with edamame. Actually. She's, she's two, um, going on three, but something about just trying to break into the, you know, those little soybeans out. She's, she loves it. And I was like, it's a healthy snack forward too. So nice.

Speaker B: Get her some wasabi covered ones. It'll be great.

Speaker A: Yeah, some wasabi with it. M. All right, good stuff. I'll uh, I'll jump in. Um, my pairing this week is recruiting and catfishing. Uh, just, it's top of mind lately. It's, it's been obviously covered in the media. Uh, we see it on a daily basis. And we actually just had two kind of client calls this week that were just kind of highlighting the problem of, uh, the huge uptick in fake candidate profiles of folks applying to jobs. Just straight up fake resumes, usually AI generated, um, and uh, most of them, you know, for 100% remote jobs, but we probably screened five to 10 of them on a weekly basis. Um, but you know, heard some horror stories from some of these companies. One going as far as having people showing up to like an in person interview and then getting the job offer and then a completely different person showing up for day one on their start date, which is, I mean that's like some, some Hollywood there. So anyways, I, I um, I feel like it's, it's one. I want to create like an entire episode around this kind of phenomenon that's just getting worse and worse. But for now I'm just gonna go with it as my, my pairing. So. Recruiting and cat fishing.

Speaker C: I've heard that what people are doing now, it's really like, put your hand in front of your face.

Speaker D: Ah.

Speaker C: They're like, well, I can't do that. Why would I do that? And it's like, because you're a bot.

Speaker A: Yeah, it's wild. There's a, there's a lot of little hacks and uh, that we've kind of picked up on, on how to get, get folks to kind of get scared off if it's like, um, even if it's a full remote role, like uh, you know, there, there could be the chance that you need to come on site for the onboarding, you know, phase. And then usually they'll be like, I'm not, not really interested anymore or something. So, um. But uh, yeah, we'll craft a full episode on that when we talk, uh, for hours on it. But let's pass it to our guest, Nathan. Quick intro and your pairing.

Speaker D: Sure.

Speaker A: Nathan.

Speaker C: Uh, my, my pairing is something that's happened recently and I think I'm gonna feel like a bad parent saying it out loud. Talkative five year old. And sometimes when I'm done answering the question why, I will give him chat. GPT to talk to. Talk to the voice on chat sometimes. And he will wander off for like 20 minutes talking to this, you know, whatever it is. And uh, I feel bad because like afterwards you can he he wants his conversation private, but of course, afterwards you can actually look at what he's saying. It's actually quite far right now.

Speaker A: That's great.

Speaker B: That is great.

Speaker D: Your answer, that is five year olds talking to chat, GPT hearing.

Speaker C: Yeah.

Speaker B: Um, that's another my go to move I stole from my wife was, uh, why do you think?

Speaker A: Yeah, I would love to see the feed on some of those questions, Nathan. I feel like that would be comical.

Speaker B: Yeah.

Speaker A: Cool. Well, thanks again for joining us, Mike. Uh, Mansell, uh, quick, uh, intro on your pairing.

Speaker D: Yeah, thanks for having us. Mike Mansell. I'm a principal with AmFam Ventures. We're an early stage, uh, venture capital firm focused on the future of insurance. Um, my pairing, I guess it's top of mind because you mentioned you were recently at Lambeau Field. Um, so I'm thinking the fall season and football is a good pairing. You know, I'm based in Madison, Wisconsin. Summers, uh, are, are too, too quick here. So fall gives you something, uh, to look forward to by way of football.

Speaker A: Nice. Good stuff. I didn't know if you're gonna sprinkle cheese curds in there, uh, in lieu of the, uh, Wisconsin. I mean, that was like every, Every person week. Did you try the cheese curds? I'm like, oh, I plan on it. But yeah, football in the fall season, it's, um, unfortunate, you know, just being a Commanders fan. Yeah. So it's a horrible, horrible season for us, but seems like the packers are doing pretty well this year. So.

Speaker D: Good.

Speaker A: Good for you guys. Um, cool. All right, I'm gonna just kind of transition us into the heart of today's discussion. So, you know, we want to cover a few different areas on this one. You know, first, kind of like defining the problem set within specialty insurance, um, and how, you know, the startup Ledgebrook is kind of tackling the problem here. More about designing for brokers versus, you know, pure insurers and highlighting where AI fits into this, of course. Uh, and then along the way, you know, ensuring that we catch, uh, the investor perspective on all of this with the viewpoint from Mike at American Family Ventures. So, uh, let's dive in, Nathan. Let's kind of maybe set the stage here. You know, what's the core inefficiency that you all see in commercial ENS insurance that Ledgebrook is solving, uh, and why it's such a pain point. And, and I'll just also, you know, say for some of our listeners who maybe aren't well versed in this kind of insurance, maybe just highlighting, you Know who ENS insurance is typically designed for?

Speaker C: Sure. So ens, uh, is designed for people that were in the traditional insurance market but they've been kicked out for whatever reason. It could be they had a ton of claims and nobody wants to insure them. It could be that the insurance company just decided, hey, I'm no longer writing any more houses in California. Um, and so what we do. So the problem space is let's say that you own ten, uh, hotels and you want to go get insurance. There's lots of different types of insurance. There's property, general, casualty, cyber, um, and you need it quickly. In general, if you were to do it through one of the legacy incumbents, you would be lucky if you get a response back within the first couple of weeks. And not a response that shows you what the premium is, what, what your coverage is going to be, but just a response that hey, I got your message, we'll look into it. A lot of times you don't get a response at all. So that's the primary issue that we have rallied uh, around and solved is just being quick to respond to our brokers, sometimes within minutes, um, not just with a clearance but also with a rated submission to say look for your ten hotels. Um, given the information you've given me so far, we're talking roughly $100,000 a year. And that has been a big boon to our success with our broker partners.

Speaker A: So it sounds like delays and speed is, is a huge pain point for anybody trying to pursue those, those types of, that type of insurance. Um, I guess Mike, just quickly from your side of the table, you know, what was it that initially kind of stood out to you about Ledgebrook when you first, you know, came across them and um, you know, why speed and quoting infrastructure kind of caught your attention as an investor?

Speaker D: Yeah, that's um, I remember vividly kind of the first few meetings and uh, I think the team was something that really stood out. So Ledgework's leadership team is um. Actually we could go back to your first question. A good pairing could be uh, technology talent and insurance expertise. And that's something that, that Ledgebrook uh, brought to the table early on. I think. What, what, what has been really uh, interesting with led and kind of key to their success is they're winning business uh, versus competitors for non price reasons. Um, you know, you don't want to be under bidding competition and growing uh, as a low, low cost option. Uh, you know, you could imagine an example and we all hear probably uh, like the Florida or California homeowner's market as, as an example, you know, if I started a homeowner's carrier, I could uh, I could grow really quickly if I offer prices that were substantially below the competition. Um, and that's not something Wedgebrook is doing. Um, you know, the market is pretty competitive. People and distributors will find you if you're the cheap option. Um, you know, the broker's job in insurance is actually to literally shop. Um, so one thing that Ledgebrook has done a really nice job is winning on um, speed and service as their key differentiator. Um, and Nathan and his team on the tech side obviously drive that. Um, so in many instances Ledgebrook, uh, in some ways isn't in that auction market at all. Um, in other words, brokers come to them really like the speed and service and bind, uh, as a first option.

Speaker A: Yeah, always. Just generally curious too, just from, um, we have a number of investors that kind of join with founders from their portfolio and just uh, kind of getting that uh, perspective on, you know, what kind of signals stand out as an investor thing like traction. You kind of alluded a little bit of like the founder profile, you know, sales velocity. What, what is it that you kind of, you know, specifically look for to help gain conviction when you're looking at investments?

Speaker D: Yeah, I think, you know, maybe broadly speaking it's a combination of product, market and team. You know, is. Is the product something that the market needs? Is the market massive and is the team, um, you know, impressive? And in this case, uh, Ledgebrook is going after, you know, the commercial ENS market, which is a space that has grown substantially over the last, uh, last decade really. Um, and so when we first invested, it was really just a pitch deck and an idea. But it had the necessary, uh, you know, high level ingredients. And then over time, um, we've kind of doubled down over time as Ledgebrook has. I'd say maybe put simply, they do what they say they're going to do. So they hit their growth targets. Uh, the loss performance has been really strong. They're able to launch more and more products and bring on more and more impressive people. And so it's really been, um, I don't want to say it's been an easy decision, but Ledgebrook has made it, uh, you know, it's been a joy to watch them continuously do what they say they're going to do. And that trust factor, something that really matters when we're making investments.

Speaker A: Nathan, I want to um, you know, come uh, join as the cto, right. I want to talk a little Bit about, you know, from the technology and the, maybe from the architecture of what you're building here. Given that you guys are obsessed with, you know, speed or quote speed. You know what, what just walk us through, like how you all kind of built the platform for speed. You know, what, what goes into that and anything maybe from your stack or workflow that makes this fundamentally faster than the status quo.

Speaker D: Sure.

Speaker C: Um, so one of the things that we started with was our policy, uh, administration system. So my personal philosophy on building things versus buying things is you want to build anything that's part of your core competency. Right. And for us that was speed and efficiency of the underwriter. Um, and then anything that can help from a data perspective that impacts pricing. Um, having a policy, administration, something, something that can, you know, keep track of policies once they've been issued, how much the final, uh, premium is going to be. Adding additional changes after the fact wasn't going to move the needle, but it did provide us a great starting point in terms of having a very well defined data model that we could enrich API ecosystem that we could build off of. And so from there, what we started looking at was like, okay, where are the pain points in the value chain? Right. Like we get an email from a broker, that is how they want to deliver business. Um, we have to clear it, we have to rate it, we have to send a quote back. Like, those are the three primary steps that happen. So we started with clearance. How long does it currently take? Well, it was taking 100, like 60 minutes, I think, on average for one. And so we said, okay, these are pretty structured forms that are coming in, but there's a little bit of discrepancy sometimes between the form data and what the email will say. And so we leveraged a combination of OCR and LLM and we whittled that down to about three minutes. Um, and after that we just kept picking and choosing where were the biggest pain points, what could we do? Completely automate. And so the next thing was easy. It was, you know, let's actually generate the rating, um, you know, the, the premium amount that we're going to charge customers. And so that those two things right there made it so that we can quickly, uh, you know, flip back a quote to the broker that says, look for, you know, like I said earlier, this much coverage, I'll give you 100, I can charge you $100,000 in premium. I would also be, I love my little anecdote. So we had a VP from a company, um, I'm blanking on the name right now. But he in insurance and uh, they did about a billion dollars in premium a year. And he came over in here and we were showing him a demo uh, of the system and literally all I did was click a button to generate a quote. That's it. Click a button to generate a PDF. Sounds like it's not that fancy, right? And he said wait, do that again. So I clicked the button, generated the quote and then he's like can you like change the amounts on there and generate again? So he changed the stuff and he was like oh my God. Like the place that I just came from that would have taken two days to get that from. Uh, the IT department, we're also known as the IT department still grinds my tears a little bit. But uh, so that's like uh, some of the things that we're up against from a legacy carrier. So the bar can be pretty low in order to you know, make the process better on every step of the chain, which has been a lot of fun to do.

Speaker B: I think it's interesting, I think you highlighted some of it. The uh, the, the amount of uh, not technical solutions in insurance. It's like there's a few industries that just have not like embraced the digital revolution. Um, and, and APIs and other things and things are still, you know it's all uh, there's a, it's all transacted over email which is definitely a, a challenge I imagine. And uh, being able to process those quickly, it sounds like that was like the first thing you identified as, as needing to deal with. Which is um, I think I'm curious like when you balance that out versus like hey, you know, um, versus say building an API where the broker can come in or a website where the broker can fill out a form and submit it. Why um, like how you guys came like that decision making process of let's do email.

Speaker C: Yeah, My, the general thought is, you know, it's really hard to change behaviors. Right? Like there's people in the space that have built out these digital portals and starting to try to make a uh, digital wholesale brokerage. But it's really hard to convince the brokers who are very used to just searching for the name of the person they remember and forwarding the email on and having those correspondences easily. And so after doing you know, basic product market fit, that was one of those things. They're like, well maybe we could build it that way. But even better, instead of changing behavior, you meet the customer where they are and uh, design around that and one,

Speaker D: one way that you might think about Ledgebrook is they're really tech enabling human underwriters. So they're, they're allowing the underwriters to be more productive through the use of technology and you know, ensure lot. Lots of um, dynamics and insurance are relationship driven. Um, and this particular, this particular area of the market is very relationship driven. So having the human in the loop is something that um, seems to matter greatly,

Speaker B: which makes sense. It's a regulated licensed space. There's a lot of. You always need the human. I mean the broker, the agent is an important part of that relationship and is you know, there to sort of help mediate that. So it makes sense. It would be a, there's always going to be a human in the loop.

Speaker A: Yeah. I'm just curious on like the user experience specifically for the broker. Like Nathan, how did you like in the team make sure that you were building with I guess the brokers in mind, not just like insurers. Like um. Yeah, I'm always curious on like how you settled in on the, the broker.

Speaker C: Yeah. So uh, my, what I always tell my team is like my, our customers, we build internal products for our underwriters. Our customers are the underwriters but by proxy they're the uh, wholesale brokers. Um, and so when you have that customer centric focus, right. It's once you get around the uh, trying to do the behavioral change, it's like what can we do to first figure out what they're actually bothered by? And so after talking to them, it's always been like, there's a lot of times we'll send out a submission packet and we'll never hear back from the insurance company. A lot of times we'll send it out and we won't hear back for four weeks. And it was due three weeks ago. Um, and so that's how you're able to identify very quickly. It's like, okay, well it's a speed component and once you establish that relationship where you do what you say you're going to do as our underwriter, um, the only thing that the broker has to remember is if I send it to Ledgebrook, I'm going to get a response. It might be a no, but it's going to be quick and I'm going to know that I'm not going to waste my time on this very valuable account to uh, with them.

Speaker A: Speaking of like, you know, kind of like building as a customer first company, you know, Mike, as an investor, you know, evaluating. Well I guess, yeah. How do you evaluate whether a startup is truly kind of customer first, you know, especially in like enterprise or you know like B2B heavy spaces.

Speaker D: I think um, maybe I'll, maybe I'll try to generalize but also relate this to Leshbrook and uh, speaking to, speaking to the industry and folks in our ecosystem is probably the most kind uh, of tangible example I would give. So in, in Ledgebrook's case, they work with wholesale brokers. Wholesale uh, brokers end up working with retail brokers. So broker feedback is something that, that we can get through our ecosystem quite easily. Uh, uh, kind of the other end of the, the value chain. Ledgebrook works with capacity partners or reinsurance companies. And we can talk to reinsurance companies and see, you know, is the business Ledgebrook writing, um, you know, performing on plan, uh, you know insurance is a, you know, in many ways it's massive. In other ways it's, it's close and small. So we can also talk to competitors and say are you seeing Ledgebrook in market? What, what are you hearing? What, what do you think of them? Um, so you know, I'd say the you know, customer and ecosystem feedback is, is a useful kind uh, of tool to use for uh, you know, when, when we invest there might not be traction uh by way of revenue or it might be super early. But there's uh, certainly signals that we can get from industry um, before investing

Speaker A: and just generally curious when you know, when do you, when does American family usually come in to an investment? What stage?

Speaker D: Yeah, we're um, we have quite a bit of flexibility. We're typically on the earlier uh side. So think seed to series B, um check sizes typically in the kind of uh, 1 million to maybe 15 million uh range. Um so we're typically early, typically kind of pre product market fit often uh, kind of sub, uh sub 100 million, uh revenue, um type investments.

Speaker A: Cool. So yeah, we want to kind of touch a little bit on AI, uh and kind of see where AI fits into kind of like what you alluded to and like empowering the underwriting side of things. So um, Nathan, can you share a little bit about how ledgebrook is. Yeah. Is using you know, AI and its workflows and you know, maybe just where it, where does it add value and where do you maybe intentionally not kind of automate things?

Speaker C: Uh, it's everybody's favorite topic right now. Isn't it

Speaker A: Hard to not be?

Speaker C: Yeah. Um, so where I have found a lot of success with it thus far is workflow automations. Um, so an example that I Got two. So one is we going back to getting back to the broker quickly. Right? There's the quote step, but there's also in between that, that's what we call the indication. It's getting back to them and saying, you know, this is roughly based off the information you've given me where we can play ball from a pricing perspective. Does that work with your client? And they can quickly say yes or no. And so we um, using AI and actual real machine learning, in addition to an LLM, we made a um, combination of basically it's a chat bot, but it can talk with our rating API and so our underwriters can say, hey, this uh, email just came in, rate it for me, is it an appetite? And it'll go check our underwriting guidelines and I'll say yes, it's an appetite. It'll go look for which um, industry, which class code should we use to actually rate it? And then it will pull the information out of the uh, the email or the cord forms and say, you know, return back, you know, roughly. It's going to be, you know, $50,000 assuming the clean loss history and, and a couple other things. Um, so that one has been great for the underwriter because it allows them to spend more time building out those relationships and less time doing a lot of the manual data entry that they would be doing otherwise. Um, and another one that's very similar but a lot more technically complex is um, loss runs. So every uh, uh, client, not a wholesale broker, but the actual insured, um, has to provide five years of loss history for us to be able to do uh, an effective pricing. So but they come in so many different forms, so many different file types. Even within the same file type, they're all different wordings, different table formats, um, very complex space. And so we've been able to again leverage a combination of OCR, real machine learning and LLMs to extract that data and put it in a form. Not just that helps our underwriters, but also helps our actuaries. Because this is a huge treasure trove of data that we're going, that we use for pricing, uh, and prediction. So those two have been fun. Where it's not good, in my personal opinion, is um, what investors want to see and what our reinsurance partners necessarily want to see reinsurers want to see more of. Are you not ceding control to an AI bot? Right, that's not going to stick. Well, um, they naturally can be a little more risk averse and having that just doesn't really align with what ledgebrook likes to do, which is, you know, having a human involved in every part of it, um, but also them, uh, it's a no go. And so. And for me too, right. Like, trying to figure out how to have a fully, you know, underwriter be replaced by, uh, you know, LLMs in its current form. Just, I just don't see it. And so it's not about replacing anybody. It's about augmenting what the individual can do.

Speaker A: Yeah, we see this all the time in critical infrastructure verticals that we get a lot of companies, um, or guests on the podcast that uh, are innovating in or defense or national security. And you've got automated systems that are being deployed on the battlefield and they're never really fully comfortable with just giving full control to, you know, that autonomous system versus having a human in the loop. I think the same applies into, you know, your industry as well, where it's like, you know, there's a lot of money on the line here and just having, you know, the trust that just, yeah, the, the bot or the AI is going to just automatically, you know, spit it out. Let's just trust it. Um, I think that carries over into so many different industries. So I appreciate you kind of pointing that out. Just a random question. Does your chatbot, uh, have a name? Everybody's got this like, you know, kind of aggressive name.

Speaker C: It was originally called Lilbot L I L B O T. But then the joke became that I was so obsessed with it that it was basically my girlfriend. And so it was renamed to Lilybot. And so Lily Bot.

Speaker A: Lily Bot. Okay, I like it. Um, Mike, I wanted to ask you, you know, you've obviously seen a huge wave of AI insuretech pitches, uh, over the last year. So, you know, what is it that kind of cuts through the noise for you or you know, what kind of sends off like a real signal that this is really something here?

Speaker D: Uh, that's actually a, uh, pretty tricky question. It seems like every company that I'm talking to now is an AI company to do with AI, you know, AI in the name, AI in the domain, uh, address. And I would say 12 to 18 months ago, um, maybe the traction hadn't caught up to the hype or maybe, um, the tools that were being put out there weren't being, uh, taken, um, as, uh, meaningfully as they should have been. But now we're to the point where it seems like brokers and distributors are really leaning in. Carriers, uh, are definitely devoting meaningful resources to it. Yeah. Employees, uh, at, uh, carriers and in the industry are also realizing they can't ignore it. So right now the thing that is interesting to see is we're starting to see startups with more traction. Uh the AI plus insurance pitch. 12 months ago most of them had you know, minimal traction. Some maybe had you know, 500k of ARR or maybe a million. Now we're starting to see some of those early entrants go from you know 1 to 5 million in ARR or 0 to 2 million in 8 months. Which maybe to get back to your question, traction is kind of a thing to note. And uh, the aspect that's a little bit more challenging to discern is um, kind of we talk about moats or competitive advantages and uh, in a space that's moving as rapidly as AI, um kind of imagining customer lock in and long term defensibility is a little bit uh, more difficult to see play out. Which you know I said there's been Companies go from 0 to 1 or 1 to 5 in ARR. The question will be, you know those companies that can get to 10 or 15 million in ARR, can they then get to 20 to 50 and can they stay there or will there be a new wave of entrance that enter and kind of uh, you know take, take on the startup uh, that that's quickly become uh, become uh, old for this technology.

Speaker B: Yeah, I mean I always think of it as like there's pioneers and then there's settlers. Frequently the pioneers get overtaken by the settlers. Uh you know the, the first ones out there have to deal with all the, the arrows and slings. Uh and then uh, and then the next wave is that that second wave of people who like can learn from that first wave. So it's, I think it's in this space it'll be interesting to see how things play out.

Speaker A: Are there other pockets of insurance that you would say are kind of ripe for you know kind of like this, like a platform style innovation?

Speaker D: Um one area that I've been spending time thinking a little bit about and my team has as well as uh, the TPA space, uh so building kind of uh new tech stacks that focus on claims and claims adjusting. Um, that's a space that uh, obviously servicing and claims is a huge aspect of the uh, value that insurance provides. And there's lots of manual process, lots of paper, lots of uh, text to interpret which um, AI systems are increasingly good at. There's also lots of people involved um, which the claims professionals are very good at their craft. Um so I wouldn't say replacing Them necessarily. But similar to how Ledgebrook Tech enables underwriters, there's lots of opportunity to tech enable uh, claims professionals, um, to make them kind of better, faster and um, better at serving their end user.

Speaker A: Very cool. Uh, I want to talk a little bit about the culture um, at ah, Ledgebrook for a second. Nathan. So kind of leading the charge there. As a cto, how do you foster this, an engineering culture that's comfortable, kind of like moving fast, but also in a, you know, in a heavy regulated space like insurance.

Speaker C: Yeah. Uh, before I answer that question, I did realize from this last section the most important thing that I think we can do as a company is change our name from ledger.com to ledgebrook AI.

Speaker A: AI. See if it's available.

Speaker C: Yeah. Um, so you know, culture wise and being a true engineering company, I think the unique thing about Ledgebrook thus far has been that everybody that we've hired on the business side for the most part has insurance background. They know what bad is, they know what good could look like. They have the um, acumen and the years of experience to be able to really say what they need. And then everybody 35 person strong on the engineering team now hasn't come from insurance. They've seen what good looks like in fintech, in health tech and you name it, we've been there. Um, and so it's not only that but as a remote company you're able to meld these groups together in various slack, uh, channels and get the instant feedback of your customer using the product. So I think that has been very good for uh, the engineering team in general. That fast feedback loop that just keeps compounding as we build these things and build out new systems. Um, in terms of maintaining the engineering discipline, I think it falls out of what I just said too. We have come from industries and companies that have embraced a lot of the best practices, um, in tech for the past few decades. Um, and so we're able to take that knowledge and even though the underlying domain has changed, the uh, infrastructure and mechanics that we'll use are still very homogenous.

Speaker A: Yeah, it's interesting. We talk to a lot of startups that are in health tech as an example. Right. And heavy compliance environment, very sensitive data. Um, and we always ask, are you looking for folks that have this very specific health healthcare experience? And it's not necessarily that, but it is, you know, folks that have worked in, you know, very sensitive, you know, data sets, uh, if that be in fintech, if it be within a, you know, national security environment, uh, but still Kind of like building with that mindset but not necessarily having to be from that specific domain. So it sounds like that kind of carries over too with uh, with Ledgebrook where it's, you know, we're not uh, looking for engineers specifically coming from insurance, but you know, having folks that are coming from some other sort of regulated industries that has some carryover.

Speaker C: Yeah, definitely be helpful. And like the primary thing, honestly that we look for is somebody with a builder's mindset, right. That really thinks about the product, um, you know, versus just. I need to get this commit out.

Speaker A: Yeah.

Speaker D: One interesting anecdote that I've seen and I've spent uh, quite a bit of time with the ledgerbrook team over the last few years and is um, you know, a curious mindset as well. So, so I've, I've spoken to, to people on Nathan's team who uh, you know, they don't have an insurance background. Uh, you know, they're an engineer, but they want to take an actuary, uh, you know, class or get an actuarial designation. And I think that speaks to the, you know, Nathan, I think you use the term builder's mindset, but, but they're curious and interested in insurance and want to all in on Ledgebrook mindset that they've been able to cultivate.

Speaker C: Yeah, good point.

Speaker A: So let's talk just about where the platform's headed. What is it that you guys have on the roadmap in terms of maybe expanding into new product lines or um, geographies or doubling down on infrastructure? What is it that you would love to see on the, the kind of the vision board over the next, you know, six to 12 months?

Speaker C: For me, um, we're always launching new products. Right. So we've been general casualty. We've had different iterations of that that, you know, we will get the notification that, hey, we just signed this deal. It's going to go. We needed to go live in uh, you know, two weeks and after the initial. Oh, the uh, you know, we were able to hit those goals. So, you know, the next year. Well, we're about to launch, um, our cyber offering in the beginning of next year and we're looking at doing, um, property maybe depending on if we get the right team involved. And I say team underwriting team that's interested in building out something new. Um, so those are always on our mind. Uh, in addition to that, going back to what Mike was saying about TPA being right for disruption, one thing that doesn't really exist in the underwriting workbench Insurance tech ecosystem is something that combines both the underwriting workbench and the claims system. So we're going to embark on that beginning of next year as well. Um, it's crazy to me when you talk to industry veterans, but they'll be like as an underwriter request from your own claims department, the losses that have transpired over the course of however many years you've had that policy and they can take two or three weeks to get back to you. So it's like, okay, let's just make this stuff real time like we've done with everything else. Um, so I'm excited about that, uh, you know, embracing the AI LLM. The two things that we're doing with that one is we are uh, what I like to call workforce AI. And so, so we've partnered uh, with Amazon and we're using their Amazon Q Enterprise. And what it allows our team to do, our non tech team, is use natural language to build out workflows that are causing them lots of pain. And so we have uh, two things that have already been built by uh, the 10 pilot users that we have. Uh, and that's what I'm really excited about to see those in production. Right. Because right now those requests come to our team and it is a big backlog of those requests. And when you're able to provide them the infrastructure and the guardrails to do it, not only do you enable, uh, them to move faster and move quicker, but also enables us to move faster and m quicker. So I'm excited about that. And then last but not least, um, we have the initial chatbot, but actually being able to converse with our data, actually being able, able to um, review policy, uh, document as it comes in real time and being able to do all the things that I know are

Speaker D: possible

Speaker C: with LLMs at this point is something that we've set for the first half of next year as well.

Speaker A: Cool. Yeah, it's exciting. Um, kind of some milestones that you have listed there and just maybe some quick hits on the company for our listeners. Um, what's the head count? I think you mentioned it's fully remote. Um, but what's the head count at? Maybe some, some info on the funding to date and then would love to hear about some of the roles that you guys are looking to hire up for as well.

Speaker C: Sure. Um, so yes, we are remote always. I what I always tell people when they're, they're um, interviewing and they ask that are you always going to be remote? And I say, well, I can't Promise that. But what I can promise is I will be the first person to tender my resign. So I think we're always going to be a remote company for a variety of reasons. Um, we have ah, roughly close to 200, uh, full time now. The engineering, ah, group I think is the 35, I think somewhere around there. Um, and we're hiring for a lot of roles. So engineering team, we've got head of it, something we've been putting off for a long time. Um, we've got full stack roles, you know, mid all the way up to senior. We've got engineering manager roles, we've got data engineering roles, we've got uh, I'm probably forgetting something, but product, uh, manager role, um, and then that's just within engineering. So even across, uh, underwriting operations, you name it, we're hiring.

Speaker A: Yeah, that's, that's uh. I don't think I realized that you guys were 200. So how, how many underwriters do you actually have on the team then?

Speaker C: We were just in Las Vegas for the underwriting summit and there were, I think 78.

Speaker A: Wow, that's awesome. Yeah, definitely is a testament to the, yeah. The quality that you place on like that call, like that White glove experience. Right. Like, you know, not just throwing things into the tech and, and uh, hoping it spits out the right thing, but making sure it's kind of paired with, with a human, um, you know, an expert in that space. Um. Cool. Very, very, very exciting. Um, I take it no. Yeah. Not looking for any sort of catfish, you know, applying to.

Speaker C: I kind of want to have the experience just so I can talk about it, you know.

Speaker B: Oh, I can forward you some resumes. I uh, I've definitely done those interviews, uh, where I saw somebody like walking around in the background and it's like there looked like a towel around. They were calling from like some sort of call center dorm room craziness. It was just as like within two seconds. Yeah, we're done.

Speaker C: Listen, that's what I mean. Give me a couple so I can just experience it.

Speaker A: Just put them on with, uh, with Lily. Let Lily do the. Let Lily do the screening.

Speaker B: There you go.

Speaker A: Um, excellent. Well, uh, yeah, we'll make sure to, to shout out some of the, uh, yeah, some of those roles that you're hiring for when we, when we post, uh, the episode. Um, outside of that, um, yeah, I think we can kind of segue into the, the final, uh, portion of the episode which is just kind of a fun rapid fire Q A. Uh, it's called the five second scramble. You know, try to give us your best answer within five seconds. You won't get fully air horned off if you don't. Uh, but try, uh, to keep it moving. First thing that comes to mind, type of thing. Uh, Mike. Why don't you start with Mike and then I will, uh, wrap with Nathan.

Speaker B: Sounds good. You, uh, ready, Mike?

Speaker D: Sure. I'm curious. Is the, the wheel in, uh, in the background they're gonna be in play?

Speaker A: Nah, we phased it out. But you know, I can give it a spin while you're waiting. It's almost like a little, uh. Let's go, Mike.

Speaker B: So. There we go. So, uh, so, uh, what makes American family ventures unique?

Speaker D: The people.

Speaker B: What's one thing about that?

Speaker D: The rapid fire you want is just short, short, blunt answers.

Speaker B: Yeah, yeah, yeah. Or I mean, yeah, you can expand

Speaker D: a little bit, but yeah, the people are great. And we've been at this for, ah, over a decade.

Speaker B: Awesome. Uh, what's one thing about insuretech that you think would surprise people?

Speaker D: Um, how vast and broad it can get.

Speaker B: Uh, what's a common mistake early stage companies make when seeking investment?

Speaker D: Not, um, not being able to succinctly describe what their company does.

Speaker A: That's a problem.

Speaker B: That's amazing. Um, where do you see AI having the biggest impact in commercial insurance?

Speaker D: Providing, um, efficiency gain to humans?

Speaker B: Uh, what advice, uh, have you received that you found to be truer over the course of your career?

Speaker D: Relationships matter and, uh, build a posse, uh, of people that are, are that have your back.

Speaker C: Nice.

Speaker B: Uh, what was your first car?

Speaker D: A Subaru Outback.

Speaker A: Nice.

Speaker B: Very nice. As a kid, what did you want to be when you grew up?

Speaker D: Probably a professional hockey player.

Speaker B: Oh, really?

Speaker A: You kind of have a hockey name. Mike Mansell. Sounds like a hockey name.

Speaker B: Yeah, it could be a hockey name. Yeah. You have a favorite team?

Speaker D: Uh, no, I'm m not, I'm not a huge NHL guy. More college hockey. Wisconsin pledgers, of course.

Speaker A: Cool.

Speaker B: Uh, what's, uh, your go to movie theater snack? Popcorn.

Speaker D: Uh, but I don't remember the last time I was at a movie.

Speaker B: Uh, what's a surprising hobby of yours?

Speaker D: I kind of gave it away with hockey. I play hockey once a week still.

Speaker B: Oh, nice. Uh, and then last one. Uh, what's a charity or cause that's near and dear to you?

Speaker D: Uh, I would say working, uh, with college students. I like to kind of mentor. I don't know if that's a charity, but call it a cause. And I get energy from being around college, uh, students Interested in insurance.

Speaker A: Nice. Nice. Good stuff. All right, Nathan, you ready?

Speaker C: Let's do it.

Speaker A: I'm gonna spin the wheel.

Speaker C: Thank you.

Speaker A: It means nothing. Um, all right. If Fledgebrook. Fledgebrook, uh, were a band, uh, what genre would you play?

Speaker C: Oh, I mean, I think it has to be classic rock. Maybe. Maybe there's a little. Little techno in there to be the, uh, new engineer.

Speaker A: I'm thinking of like a Bob Seeger. I dig it.

Speaker B: Yeah.

Speaker A: What's one non technical skill every engineer should have?

Speaker C: Oh, communication. Just being able to converse with people.

Speaker D: Yeah.

Speaker A: What's something about the culture at Ledgebrook that you would want new employees to know about?

Speaker C: Uh, we love to make fun of each other, which has been a lot of fun.

Speaker A: That's great. Uh, what's the most surprising skill that you've had to learn in your career?

Speaker C: Uh, tech support at this job when I first started.

Speaker B: Nice.

Speaker A: Need that director of it. Um, if you weren't building in Insurtech, what other industry would you be tackling?

Speaker C: Oh, man, I don't know. Um, you know, I mean, how. Trying to think of something that hasn't been tackled already. But the only thing that's coming to mind is going back to AI. I hate myself for saying it.

Speaker A: Uh, what's your favorite travel destination? For personal travel?

Speaker C: Oh, my favorite place I've ever been. Probably Hawaii.

Speaker A: Nice. What was your very first job?

Speaker C: Very first job? I, uh, actually taught. Got, uh, inner city kids programming when I was 15.

Speaker A: Wow. It's cool. What was the name of the. Was it just, like a nonprofit or.

Speaker C: It was through the Southwestern Athletic Conference in Birmingham, where I was from.

Speaker A: Nice. Cool. If you had to teach a master class, uh, on something that has nothing to do with your job, what would it be?

Speaker C: I. I love writing parody, uh, songs, so that's probably what I would go with.

Speaker A: Oh, cool. First time we've ever heard that. So it's a unique points. Favorite pizza topping,

Speaker C: sausage.

Speaker A: And then last one. What's a charity or corporate philanthropy that's near and dear to you?

Speaker C: Uh, Alzheimer's. So I've got some support that in the family.

Speaker A: Very good. Yeah, we'll plug both of, uh, both of those in the. In the show notes as well. And the episode gets pushed live. But Nathan and Mike, thank you guys both for joining us on the Pair Pro program. Really enjoyed learning about the market and excited to see what's next for Ledgebrook and American Family Ventures. Thanks for joining us on the podcast.

Speaker D: Thank you.

Speaker C: Thanks.