a16z Podcast - AI for America's Small Businesses | Lassie

Episode Date: July 30, 2026

Alex Rampell and Olivia Moore speak with Lassie cofounders Steijn Pelle and Frédéric Renken about bringing AI to one of the most overlooked parts of the economy: small businesses. Inspired by time spent working inside dental practices, Pelle and Renken set out to automate the administrative work that keeps healthcare providers away from patients. They discuss how AI agents are changing billing, insurance claims, patient payments, and other operational workflows, allowing practices to spend less time on paperwork and more time delivering care. The conversation explores AI agents, software that performs work rather than simply storing information, onboarding AI into real-world businesses, and why healthcare administration offers one of the biggest opportunities for automation. Along the way, they discuss product design, go-to-market strategy, and what it takes to build AI systems that operate reliably in complex business environments.   Resources: Follow Steijn Pelle on X: https://x.com/steijnpelle Follow Frédéric Renken on X: https://x.com/fredericrenken Follow Alex Rampell on X: https://x.com/arampell Follow Olivia Moore on X: https://x.com/omooretweets Stay Updated:Find a16z on YouTube: YouTubeFind a16z on XFind a16z on LinkedInListen to the a16z Show on SpotifyListen to the a16z Show on Apple PodcastsFollow our host: https://twitter.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

AI is overhyped in Silicon Valley that's underhyped in Iowa. I would actually argue software just kind of took things that were stored in paper format and then they made them available first on-prem via green screen computers but people still had to do the work.

I never forgot what I saw the number one great doctor on Yelp spending 200 hours a month on paperwork. The models are trained on so much data and they're so large and yet they actually don't really know how to do any of this work. Initially we were actually the humans in the loop.

We kind of automated away our own problems. The battle between every startup and incumbent comes down to the startup gets the distribution before the incumbent gets the innovation. We come by and we say, hey, we actually have built agents that can provide you already with tens of hours of labor. They then adopted, like, very quick. They see us as someone

they break in to actually run the practice for them. There was a great quote from Dr. Kwan about you guys, which is that Lassie isn't replacing humans, but like freeing them from wearing so many hats. It's not like, oh, AI is going to take the jobs. In many cases, you can't find someone. How are you prioritizing?

What you build, who you sell to? Is there of the world where Lassie for dentists makes Lassie for physical therapists better? The end goal here is that AI isn't just changing how software is built. It's changing what software does. For decades, software mostly stored information. Today, AI can increasingly perform the work itself. In this episode, Alex Rampel and Olivia Moore speak with Lassie co-founder, Stein Pella and Frederick Brank. about building AI agents

that automate administrative work for health care practices, from insurance billing to patient payments. They discuss why small businesses may be one of AI's biggest opportunities and what it takes to build software that operates autonomously and why the future of enterprise software

may be measured less by features and more by the amount of work it can take off people's plates. So welcome and thank you for joining us. Thank you for inviting us. I'm going to be here. Maybe we'll start with the basics, So Stein, this whole company started with a conversation with you and your own dentist, Dr. Kwan.

What did he tell you that made you decide to quit your tech job at Robin Hood and go process payments for him by hand? Yeah, I did not know that my American dream would look like this. It was interesting. Like, I was at Robin Hood at the time. And I came to this country to start a company. So after six years, I moved from Amsterdam to Silicon Valley, like I was looking for a heart problem to solve. And then my doctor, Dr. Kwan, I was a patient there. So I saw him twice a year as you do with a dentist.

knew that I was looking for a heart problem. And he said, do you want to see how I run my business? And I said, absolutely. And he walked me to the back and I never forgot what I saw there. A small business owner that is the number one rated doctor on Yelp, spending 200 hours a month on paperwork and busy work. So submitting claims by hand. And he had to stick around himself because he couldn't find people to build the patients.

So I'm like, wow, it's fascinating. This is a couple of years ago. So I thought that this was a solved problem because in the 70s, my mom worked in hospital. and that's what she did. She brought bags of cash to the bank and then processed payments by hands. But it wasn't a solved problem here.

So that's why this piqued my interest. Although I have to ask, when he gave you this offer, were you like, was it that kind of reclined? Could he actually process your answer is yes versus no if your mouth was open and drills were in your mouth? Yeah, yeah. How did that go down? Well, up until now I don't have cavities.

So, you know, there was no drilling happening yet. So, you know, no con would. Examination. Exactly, yeah, yeah, yeah. Exactly, yeah. No, he took me aside after that. Okay. Because he knew that I was like looking for this heart problem and was roaming around.

And then after that appointment, he showed me kind of like what was going on. And then I thought maybe it's him, right? But it didn't really make sense to me because he was very well rated. He used all these modern technologies. So then we started talking to other doctors, like because maybe this doctor one was just an anomaly. But then I also worked for a gastronologist in Scranton, Pennsylvania. And we saw the same there. I'm like, wait, you're doing this all by hand?

So then we figured out, wait, there are like hundreds of thousands. of these small businesses that literally do this all by hands. That would be quite a fascinating problem to solve. We knew it was going to be a hard problem, but we were kind of looking for that. Yeah. The Lassie story is so unique to me because you both spent months, if not years, before kind of fully releasing the product, like literally in the office of the customers. Yeah.

How did you convince them to let you in and to kind of look through the heart of the business and get into the financials? Yeah, it's a little weird, right? It's like, hello. I work at Robin Hutzel. growth on the referral program. Can I get a job here? And by the way, Frederick worked at superhuman on products. Can we do the billing for you and take over the finances? I think that was the first sign that we were onto something big. Because to our surprise, all these doctors,

when we asked them, so we first talked to all of them, like Dr. Guam, because everybody likes to talk about their problems. And when all these doctors started talking to us for hours, we knew that, again, this is a real problem they have. This is not some vitamin that maybe it's nice to solve that for them, but this is something that keeps them up at night. It makes them almost quit their job and say, I got into this industry because of my passion and craft, in this case, because I want to take care of patients. So that was the first sign that these people were not, no chance that you can come work for me because you don't have any experience running the finances. What about HIPAA and security reasons that you have access to all this information? So I think the first sign to

us that these people said, yes, Dr. Kwan said, just come and sit here, night five. You can do the job. Dr. Shah was in Scranton, Pennsylvania. He set us down behind the desk and said, you can have access to anything you need to have access to or a teacher like my son. So that was the first sign that this was very broken and not a solved problem. And then I think that triggered our intuition for, okay, we might be on something because this is a real pain that people are desperately looking for a solution. And from a technical perspective, so Frederick, business started in 2020 and so much was different then in terms of what was even possible to build. Like, how has your product building process changed over time?

How is what you thought possible then different from what you think is possible now? I would say when we started the business, we were always obsessed with automating and putting the business on autopilot. So that hasn't really changed. Back then, the models weren't that good, though, especially reasoning models, didn't really exist in that form. But if you think about what it takes to automate any. any job, you're really looking at getting context on the work, which in the case of a doctor

office is basically you need to have access to all the historical data, the patient records, that sort of stuff. And then you need tools to do the work. You know, this is true if you're a human in the office or if you're an agent. And so we started building the context layer and building the tools. And it's just that the intelligence layer wasn't that intelligent. But for the first job, it wasn't that necessary. Like the most basic kind of automation didn't require that much reasoning. But then as the models got really good, we had this kind of huge tailwind because we had all this context already built, all the tools already built, and we could kind of, as the models got better, just replace our intelligence and the product would just get smarter over time. So in that

way, we got a little lucky, but I think the core vision hasn't really changed at all. It was always about automating the work and not building tools for them that they would have to use. I feel like as the models have improved, we are more and more seeing software do the job of labor, which Alex, I would say to the first to argue and famously argue that would be the case. Curious how you think about that when you look at companies and maybe how it played into like the thesis around Lassie. Yeah, well, so I've given this whole presentation on the origin of software was basically take a filing cabinet and put it into a database

and kind of pick the time equals zero moment for that with this company, the Sabre Systems, because airlines would just keep reservations and filing cabinets. Sabor Systems was a joint project between IBM in American Airlines, that's why Sabre is spelled with 2A's. Sabre. But then this kind of took wind everywhere else. So, like, there are HR filing cabinets, and that became something like PeopleSoft. There are legal filing cabinets, and that became all of these Lexus Nexus products. Their accounting filing cabinets, and that became QuickBooks, and that became NetSuite.

So that was the origin of software. So software just kind of took things that were stored in paper format, and then they made them available first on-prem via green-screen computers, because that was a lot more to book an airline ticket and change it if you didn't have to use an eraser anymore, starting with Saber. But people still had to do the work. So I would actually argue that the world didn't get that much more efficient with software, because all that software did was, like, take HR, like, did people soft and then workday make HR departments more efficient? Like, I don't think so because the same number of people worked in HR for the exact same

size company in 1950 as probably 2000. And instead of using filing, cabinets that are guarded by Stein and Frederick, making sure that nobody breaks into the HR files. Now you have an IT department and a C-SO to make sure nobody hacks into the IT files or the HR filing cabinet. So nothing really got more efficient. I'm somewhat exaggerating for effect here. But what you can now do with software is it can edit the filing cabinet, right?

It's no longer just the dumb storage. It's actually like the smart implementation of changes against those things. So if it's HR, let's do a background check. Or let's do an onboarding, or let's explain the benefits to this person. If it's accounting, what do you do with the financial statements? I imagine I'm a dentist and I see I have all these overdue invoices and I can look up in QuickBooks. What do I do?

Well, I might want to call and say, please pay me. Like, that's what the filing cabinet should be doing, not just giving you the information. So it just turns out that the work is orders of magnitude bigger than the storage of information that the work is done on. So that's been the thesis. And you need the technology to catch up. can actually do it. Because in 2023, it's like, you know, next word prediction

wasn't really good at like going, which is basically what AI is. That was not good enough to go say, I'm going to go run my practice or do background checks. I know I'm going to have statistical inference play out, and that's how I'm going to do a background check and make sure that Frederick didn't commit any crimes before I go hire him for my company. Like, no.

But now things have gotten good enough, and that just massively expands the market size. And if you think about fintech, FinTech massively expanded the size of many non-financial markets because now you could bundle in financial products with non-financial products. And what I mean by that is, like, my favorite example of this is Toast. I know you and I have talked about this a bunch.

But, like, Toast could have existed in 1985. Like, you know, everybody had an IBM PC. They worked pretty well. Microsoft DOS worked pretty well. Why didn't, why wasn't there a restaurant software company in 1985? Well, number one, it was too hard to use, but then number two is you have this cacte LTV issue because could you get a big restaurant

that grosses $5 million a year to spend $100,000 on an MS-DOS software product for keeping reservations and paying waitstaff and having a little menu that showed up for the cook so they can make your hamburger more quickly or something, nobody would pay $100,000 for that. But if you bundle in payment processing, you're effectively charging $100,000 for that, right?

because maybe you get a 2% Vig. So Fintech made the market much, much bigger for software because of this bundling effect. And that pales in comparison to now software doing the job of laborer because it's like, yeah, FinTech made it a little bit bigger. But now, instead of just being a dumb pipe for data or a dumb storage of data, and instead of just like charging incrementally more

by bundling in financial processing, now we can do work and we can charge for work. And we can charge for work at a way is cheaper than humans, better than humans, but I think both of those sell the opportunity short because in many cases, you can't even find a human. Actually, the best and funest story of the Lassie introduction or like our announcement that we made together with you,

or your announcement, your amazing video. It was a great collaboration. So my first dentist, hopefully he's listening to this podcast, his name is Ronald Sloop. It was my parents' like first friend when they moved to Florida. I'm from Florida. He retired as a dentist. Was he a Dutch?

He sounds very Dutch. No, you know, Ashkenazi Jew for, you know, summer in Poland, Ukraine, whatever, my family is from, too. So, you know, but he's probably 75, 80 years old right now. But part of why he retired was he lost his, like, you know, key woman that did the books and everything else. He's like, I can't deal with this anymore. I quit. And then he sold his practice to his, like, junior practitioner. And now he's out of the dentistry business.

So he saw this announcement. He was like, oh, my God, this is amazing. And I'm not talking this up because you're able. He's like, my dad called me about this, because he saw the press release. Wow, I talked to Dr. Sloop about this. And he said that if this had been around, he wouldn't have retired.

Like, this is why he's now, like, doing nothing with his life of just, like, you know, playing golf or something in Southern California. He moved there from Florida, apparently. Because it's just too hard to hire the person. So it's not like, oh, AI is going to take the jobs. In many cases, you can't find somebody. This is the part that people don't realize, or you can't find somebody.

But there's, like, imagine, that there is something that every human on earth would pay a dollar for, but the cost of manufacturing that thing is $100. You just have a market failure. And I kind of call this everything to the right of the supply-demand equilibrium point on like an econ 101 graph. So it's like, well, everybody would, you know, have somebody, like, why isn't there a Dutch receptionist at every dentist in America?

Because, you know, there might be a guy that only speaks Dutch that shows up. Like, why not hire somebody who speaks Dutch? Well, because there's only a one in the hundred chance that a Stein. that only speaks Dutch shows up at their office. You're going to have to pay that person 40,000 euro. Like, it just doesn't make sense. But if it were free or if it cost a dollar, then every dental receptionist would have a Dutch counterpart, right?

You know, stuff like that. So it's just, it's anyway. So that's where the market just expands massively once you throw in labor because it's like you have this tiny, tiny market for software, which, by the way, is not that tiny. It's like a trillion dollars. Concentrically around that, you have like, you know, big on financial transactions, that's even bigger.

That's why Visa has a very big market cap. White toast can exist or... White toast can exist. But then you go, the concentric circle around that is just like, you know, it really is orders of magnitude bigger. And we also see that. So like the... There are about like 160,000 dental practices in the US alone.

And like they spend roughly $200,000 a year on like administrative costs. And then the interesting part is that, because we serve hundreds already. Like, they can't find people. So, like, what Alex, like, shared is we come across that literally every day. That, like, it is the doctor themselves with their Harvard degree that sits there till, like, midnight. And it makes them, like, not like their job anymore.

So I think that's very interesting because, like, these small business owners, they just want to mainly spend time on their patients. They don't really want to spend time on the administration part, let alone, like, working with Betty. And then in this case, it's wilder. They can't find Betty. So we combine, I think a lot of people are also surprised But isn't there a lot of skepticism?

It's like, no, these people are in real pain And they are to Alex's point about to quit Or just like they hate their job, at least this part of the job. So we come by and we say, hey, we actually have built this agent That can provide you already with tens of hours of labor And ask your friends or the people in your study club Like if this is real and then, or they Google it And they see that this is real.

They then adopt it like, very quick. So it's super interesting to see. And then indeed, like on the, why that's an interesting business, because this is indeed, it comes out of the P&L on the, on the, on the, on the, labor budget. So like we are already charging five figures for kind of like this first agent that only does 30 hours of labor a month. And there's 200 hours of laborers to be done for Dr. Sloop. And, and that's like really interesting, like to see that like they see us as someone they break in to actually run the practice for them

which is also on the other hand complicated to do because if all of a sudden the requirement for software becomes hey this is not a tool that I give Dr. Sloop and then Dr. Sloop is still on the line in fact one could argue a lot of AI companies are still like that there is a human in the loop that ultimately the software engineers

decide what's get deployed like we can do that so we needed to build an agent it's why it took us years that like errs on the side of correctness because like if you take over a job reconciling all the insurance payments, interacting with the patient to kind of like bill, it needs to like work. So that was like technically like hard to do, which also makes this super interesting from technology perspective because you all of a sudden need to build autonomous systems that run on its own and don't have a human in the loop. It runs the business for Dr.'s loop, which makes it technically super interesting.

Yeah. There was a great quote from Dr. Kwan about you guys, which is that Lassie isn't replacing humans. but like freeing them from wearing so many hats. And your launch video had a clip of him talking about how he can actually coach his kids soccer teams now and go to their games, which is amazing. There's a gap between wanting that and being willing to adopt AI and actually, you know, having it run payments in a practice and in fact doing it so well that most of your growth is word of mouth. So it's dentists recommending it to other dentists. They do. How did you approach the technical build process for that?

What was it like getting the product to? I think you guys are at 98% automation. Like walk us through kind of that journey. Yeah, I think a big part of it was us actually spending the time in offices doing the work ourselves. I don't think we could have built a product that works as well as it does if we didn't know how to do the job. I think another part, we already kind of talked about it, but I think a huge difference between SMBs in general and enterprises is that in SMBs, there's nobody to use the tools. Like you can build a tool,

but there's nobody sitting there that's going to use it. And so I think... Go for the loop at night. Yeah. Needs to go into the tool. We, from the very beginning, we focused on, you know,

initially we were actually the humans in the loop. So we kind of took over all of the work and we were like, you know, we'll just do this work for you. And we kind of automated away our own problems. And then at some point, you know, we got to, a high enough level of automation

that we felt comfortable handing it back back over to the remainder back over to the office. And I think now we learn, obviously, when we can't do something for some reason, which is pretty rare, but say we don't know how to do a certain case, we learn from what the staff tells us. And we also think about, I think, you know, like we want to get to like a sufficient level

of automation across any product to, before we sell it. So, you know, for us, I think that's like 95 plus, say, but not necessarily 100. Like, I don't think we're going to wait until we get into 100 with one product and then do the next one. I think the, we really think about the business more like as a whole, like how much of the business can we automate

and how much of the labor can we do with software. And as soon as we can take over a job, we take it over, and then we move over to the next one. And then over time, we'll just learn the long tail of cases. for a business it also doesn't matter that much right like if um let's make dr sloop famous in this podcast he's going to look yeah if dr sloop um said he can become a customer so maybe we should talk about like someone else to still activate him yeah yeah he might he might come back that's how it's yeah yeah dang that's the series b story we've got dr mack out of a stoop out of retirement we had a dental

shortest and now we don't exactly said come out of retirement yeah we got 100 million more people in the States that get good dental care. It's fine for like a business owner if there is a tiny sliver left of claims that need to be touched every week, right? So the way to see this because they live in a nightmare world where you have like to update 200 ledgers a week because that's how much patients you see and you have to like go to an insurance portal, update like a system of record check against the bank account 200 times a week. If you instead need to do that like a handful of times rather than 200 times, it saves 10, 20 hours.

if there are a handful of claims that you need to file yourself, but the majority kind of like is on autopilot makes it tremendously more easy to run a business. I often compare it with like how we are served as tech companies. It's the third company in building and it is a lot easier. And it's with a lot smaller team than we used to do kind of like 10, 15 years ago. And it is because there's a great tool for us that we can use. Are all these tools completely like running our finances autonomously yet, or our payroll or HR, no,

but they do save a tremendous like amount of time. So I think that's how we approach like this, like built as well, that we didn't want a human in the loop because we want software that skills and can be implemented quick. But it's fine if it does in this case for the first agent 98% of the work, and then there's a sliver left. And the interesting thing, we have thousands of staffers that are basically giving us input

on how to make that appeal that the agent currently cannot do. We come from the consumer world, right? Like Frederick worked at superhuman. I worked at Robin Hood. So we have a very high bar for shipping stuff. So we don't release it before. It actually like, okay, this is good. It can be hands off and it works.

And then these like staffers help us to kind of like get it to like an even higher percentage. Yeah. From an implementation and onboarding perspective, you guys integrate with existing practice management systems for the most part versus kind of making them switch a bunch of software to adopt Lassie. And Alex, you have written. and talked a lot about kind of startups getting distribution before an incumbent can innovate. Curious your thoughts on like that in the AI era and then would also love to hear from you guys

how you thought about which path to take there. Yeah, I mean, I had this epiphany when I was building my company, which is holy crap, like there really aren't, if you build something, I call this the TiVo problem. And TiVo famously, TiVo and replay TV both invented the digital video recorder so you could pause online television, which is an amazing innovation. but a terrible company, because you really have very few outcomes that are good. You either end up selling

to one of the big guys like a Comcast or Time Warner cable, but they're not going to pay you that much. Partially because if Comcast bought you, you start TiVo, Comcast buys you. Well, all of the competitors to Comcast, they're like, well, we're going to not allow this to work. So, like, you have what I'd call a control discount

versus a control premium. So that's option one. Option two is they copy what you've done. Many years later, much crappily err, that's a word, because they have all the customers. Or, you know, maybe number three, you do a licensing deal with them, and they take all the economics

because they have all the customers. So hence, you know, my recognition was, like, the thing that a lot of startups should do is they should do the boring thing. Like, they should build, like, the raw pipes, the raw, like, just own the customer, and then you get to build the fun feature on top. which I still stand by.

I mean, I like the vast, and this is why like maybe four years into trial pay, I realized what I should build is this thing called Stripe, and this was not like revisionist history, because Stripe had five people. It's like, wow, we should do boring payment processing, which is a commodity business, because if we do that, then we own the customer.

Chronologically, you get them first. It's a very, very boring thing, but then we have this other thing, which in my case was offer-based payments, which was very lucrative. But you can only do that if you control the pipe, just in the same way that you can only build the digital video recorder if you have digital video to record.

So how does this change with AI? It changes with AI because in many cases, these are non-categories. Like there is no incumbent. So for most categories, like imagine that I say, I have a great idea. I'm going to do background checks for new employees as part of onboarding,

and I'm going to integrate with Workday, workday being the biggest HR information system. That's a great idea. However, it's such a great idea that it's a very obviously great idea that this thing called Workday might copy and they own all the customers and that's where it's like,

you know, battle between startup and incumbent, it's like the incumbent might win there because their ability to add things and I feel like that is actually magnified in the AI era because you can have, like, why are big companies not good at replicating small companies?

There are lots of different reasons, but the ones that, one of the reasons is they hire very bad engineers and they have lots of process, but now AI kind of makes a bad engineer into like a pretty good engineer, you know, kind of. So that excuse kind of goes away a little bit. But this is the cool thing about a lot of the AI software companies

or the AI that does the work. Like who is the giant ass incumbent of dental software? You and I know the answer on this. But it's not the same thing as it's like, ooh, here's workday. It's a tech company. They already have software and they can add something to... I remember actually, this is a cool story. There was a company, I think it was called X1,

Microsoft Outlook had really bad search. And this company that was funded by Idea Lab and all these VCs, you know what they did? It was like search for your Outlook email, which was so good, but it's like, you know who I think is going to do this eventually? I think, and again, like that company, unfortunately, went to zero, or actually Yahoo bought that a long time ago.

So I kind of think the same rules apply. You know, the battle between every startup and incumbent comes down to whether the startup gets the distribution before the incumbent gets the innovation. But with many changes, One change is the incumbent can get the innovation much more quickly. But the other is that there are a lot of categories where there never was an incumbent software company

because the only job to be done was like actual human labor. And that's really exciting because now you don't have to worry about like, oh, shoot, these guys are going to come in and eat my lunch. Like who? Right. Like, who does? Like, there are a lot of industries that just don't have an incumbent software solution. for the industries that do have an incumbent software solution, yeah, the risk is very high,

that they will start releasing AI features. And it's a really interesting way that the market is playing out right now because if you look at the public markets, the public markets are saying in many cases, like, oh, you're a software company, software is dead, software sucks, software is zero. Oh, you're an AI, it's the opposite of what VCs are saying.

It's like, oh, my God, you do AI stuff, but AI is software. Software is AI. Like, the two are the same. But, like, I don't think everybody's come to that really, yet. So like if you're doing workday but AI, if you're doing NetSuite but AI, it's like, you know, the AI, the Pure Play

AI thing is software at its core and the Pure Play Software thing, it's like they're smoking crack or not showing up to work if they're not working on implementing AI features because that's what their customers are demanding of them. Yeah, and we see exactly that is that there isn't

an incumbent that kind of does this job or can do this job quickly. Well, the incumbent was name Betty and she quit two weeks ago. That's the incumbent. Or a billing agency. It's Dr. Sloops' old assistant. That's the incumbent.

Or a Ska-Sluppi version of that that like is somewhere overseas or in the States. So that's indeed exactly like what you're competing with. And that's what entreat us so much about these small businesses. Because like there is no major player there. And then if you show up with software that they have never seen before, which is now possible, they will adopt it and you can grow. which is very defensible. Now, maybe we get to talk about it later,

the schlep you have to do to do the actual labor and talk to all these like systems and you need to build an ontology to make sure that everybody in this whole ecosystem is on the same page about an insurance claim and the patient payment because all these different systems have like a slightly different definition of that,

which makes it also then harder to build because there is no incumbent. You need to kind of like stitch like a lot of things together, but it makes it extra defensive. because what we had to do was like, okay, first figure out kind of like all these read and write integrations between all these systems that Betty, the AI version and need access to. Then you need to figure out like, okay, what is the data model that can be used across like all these like systems? And then on top of that you need to build agents that, yeah, you can't really build without actually doing the work. So there's like, here's a work that you need to do to kind of like get that like going.

makes it very defensible. Like the go-to-market site as well, right? It's just like you need to knock on millions of doors and say, and the interesting thing we just talked about it is not necessarily that are skeptical about AI because Dr. Sloop is like, oh, I wish this was there. It's how do you get hold of Dr. Sloop, right? Because Dr. Sloup is not done at 7 p.m.

There's an emergency patient that calls. Then he goes back into the office, treats that patient, dinner with kids, and then opens the computer. And then, oops, the suppliers need to be ordered because Betty left, so I need to do this myself right now. So like for us, but for any company selling to SMBs, the interesting puzzle here, that's why this is also super interesting go-to-market work,

because like how do you adopt and spread AI in small businesses is a super interesting puzzle to solve? But that's the crux there. It is like, how do you get a bit of busy, non-technical owners that adopt AI? Well, I imagine the other part of the question for you. I'd love to hear your thoughts on this, or I'm sure our audience, but I love to hear our thoughts on this. But this is not like you download the AI app from the App Store and then you're all done.

Right? Like how do you actually do the onboarding? Yeah. And how much of that can you automate? Because that's part of what makes a business that is selling these things work or not work. Right? Because if you have to send your own Betty to every single office in the country,

that's really hard, it would be amazing if it's just like download the Lassie app from the app store and then you're all done. But there's a lot of, like, these systems and processes are very manual. And I kind of think, like, AI is overhyped in Silicon Valley, but underhyped in Iowa. And, like, there are a lot of people in Iowa. Like, how do you solve that, like, glass mile distribution problem for Lassie? Yeah.

Yeah, I think it's a super interesting problem. Assume you can knock on all these doors and get them to try it, right? Then, like, how do you get this adopted? Because the same argument still stands. They're very busy. They are not technical. So, like, good luck setting up AI in their like business. And I think that's also why our consumer backgrounds come in handy here.

Because at Robin Hood or superhuman, you get the time we know 48 hours. If the thing doesn't work and you provide core product value, you're out, right? And these doctors are very much the same. It's not only already hard to reach and hard to work with, but if it doesn't work in a couple of months from now, like you're out of the door. It really needs to be plugged in and then do the job. So I think a lot of the work went into not only building this agent, but also how do you set people up on that agent, such it almost all the friction is gone.

And that was like a lot of work. And it's already to a point that it's self-serve, almost. So there are a few more things like left, but it's literally that the doctor in Iowa, let's not use loop again, like says, yes, I want this. They then go to almost like a stripe-like checkout or a rappelling like onboarding, like onboarding, like flow where they hook up the bank account of the practice in the product.

They link the system of record. They link all the insurance portals that claims come in from. It confirms business information. Like, are these the doctors that actually work in your practice? And then under the hood kind of like configures like things. There's like one or two pieces left. But I think we're like months out until you basically like have an agent that you can almost like set up like self surf. So I think that was a big part or is a big part of bringing this technology to people in Iowa.

is that can you almost build a consumer-like onboarding like flow where a lot of the complexities abstracted away and kind of like happens under the food. Like the, maybe not many people, I think I'm very impressed by Robin Hood. I'm a little biased because I work there. But like in order to set up an account for a user without a human in the loop,

which is what Robin Hood pioneered because back in days, there was Charles Swap and then, hello, I want to open an account before that you had to go into the office. But one of the many things that Robin Hood pioneered, is that you can almost self-serve your way onto an account and KYC is getting done, your bank is getting linked, and a lot of product work went under the hood there to make that happen. I think we ran into a lot of these similar situations.

It's like how do you connect all these insurance portals and kick off like the process such that the digital claims are coming in, like reliably. How do you reliably link a bank account, all the system of records? So I think a big part of this was figuring that piece out. Yeah. I guess to that point, like there's more you can build in our building for dental practices,

then there's all these other types of healthcare practices that could use Lassie. And then there's like the broader universe of small businesses that could use you guys. How are you prioritizing what you build, who you sell to? Is there, you think, a world where Lassie for dentists makes Lassie for a physical therapist better? The master plan.

Yes. Exactly. Are we in that part of the episodes? Yes. Yeah, I think three steps. Like the end goal here is that every small business should run itself, right? And the busy work is done by agents. We want a built agent for the business.

That then will interface with the personal agent of a consumer highly likely. That then will interface with an agent at the insurance company or other parties that the business needs to interface and interact liquid. But step one is to Alexis point, like there are 160,000 dental practices in the U.S. alone, $200,000 in labor that Dr. Sloop and others can't find. So like serving that market first, you're looking at a $1 billion in recurring revenue as a market. So that's like step one. And then likely, like we will pick another doctor office type that like is not well served and has a big damn and needs consumer like product, right? Because what we discussed,

the big part of not this, you get the AI to work 95% accurate is it needs to really work. the onboarding needs to be as simple as onboarding on Coinbase or Stripe. So it will likely be another like doctor office type. And then the last part there is I think we've then trained AI agents to like run the small business and all small businesses at an abstract level, like have a system of record they need to read and write into. They all have customers in doctor offices. They happen to be patients.

But it's interacting and transacting around payments. You need to book appointments. So I think the end goal is if we served all the doctor offices, that we help all the small businesses across the world, because we're just the best in building AI agents, that salt of the earth people or people in Iowa and Paducah, Kentucky, and hopefully in Amsterdam, like down the line where I'm from, and Germany, Hamburg where Frederick is from, can start, like, using as well.

So that's the end call. But I think, again, similar to superhuman Rubin Hood, we worked at, like, laser focus on getting one thing really, really right. and then scale it like from there. Yeah. But the end goal is to help them all. Amazing. I love that as a master plan.

It's a good one, a big one. You both have been part of scaling many important companies in the past, Robin Hood and Coinbase and superhuman among others. And building a company in 2026 is like a whole brave new world. It's so different than ever before. What are like the biggest things you've carried over? You mentioned some of them already. And then what have you kind of.

of had to unlearn or what do you think is new to being founders right now? There's a bunch of stuff that we're applying now that I learned at Superhuman. I think for one, focusing on the right ICP and being really strict about who you onboard to basically guarantee that they're going to have a great experience. I think in our case, it's particularly important because if we onboard the wrong practice and say we can't actually automate that much of their work, then now we're kind of stuck with this customer that, you know, we, we claimed we're going to automate a bunch of labor for them. We can't do it. Are we going to do it? Are we going to offboard them? It's particularly

painful, maybe more painful than in a kind of old world product. So there's that. There's also, we talked a bunch about the onboarding already, but we've really obsessed about getting to the core product value really quickly. And almost like our onboarding is a little bit like a, you know, it's kind of like a story or a playbook or like a movie. We have like set points and checkpoints that we want to reach in certain time frames and we make sure it happens every time. And, you know, and we measure that, of course. I think one thing that's really different, in particular about product building from back then to now

is that if you think about the products that you built in the past, I think it was much more about kind of functionality or like the ability for the user to do something. And now we think really only about like what kind of labor can we automate and what or like work can we do and where can we save time. So if you're thinking about, say, like patient billing as an example, I think previously you would have built, you know, the ability to send a statement and the ability to receive a patient payment.

But if you're looking at where does the actual work of patient billing go today, it's really like figuring out is the money or is the statement that we're going to send the patient the correct amount and or like once the statement is sent the patient calls and asks about like why do I owe this amount? Do I really have to pay this? I thought this would be covered.

So if you're just looking at like where does the time go it actually goes in like oftentimes the customer communication or some other kind of more like you know fuzzy part of the work and when we're thinking about like shipping say a product like that we're really thinking about okay, once we, you know, once we deliver patient billing, the office should not have to spend any more time on patient billing, which is super different to giving them a tool that they then need to use, which doesn't really save them a lot of time.

And then neither of you, I think, were dental experts before you started the company, although... That's safe to say. You go twice a year. That's pretty good, I think, for the average patient. You're supposed to, right? Yeah, of course. I'm just doing my... Not a safety duty, yeah, I was about to say. Maybe tell us, like, how big is the team now? When you're hiring, are you looking for expertise in dental? What are the kind of characteristics of team members that you want to hire?

Yeah, which is not mainly on our mind, right? Because like we found there's a really great product market fit in a large market where you can go after the labor that they can find. We currently have two takes in that, like one, not much has changed, contrary in maybe you take here. you still need people with steep slope that are very ambitious and driven

and have skills that are just like top five percentile either in engineering or in selling, right? And I think that remained the same. Getting hold of doctor slope is just like the same exercises before. You could do it in a more AI like native way but I think the skills are the same

and you're, so I think that has not changed. Maybe the only thing is that the AI builtness of a person I think we see a pretty clear the vision between across the board, like we believe that the way you code will change completely as a result of building a company in this era, building out of finance department,

do you think that's going to be completely different than before? So in that way, we interview specifically for that, because we want to build a 2026 version of a big organization, right? Where like we ship twice as much than others, we move like four times as like fast, because we should not only have AI adopted in these businesses, but the big puzzle for us is how do you build a team that kind of like also incorporates AI in all these like functions.

So I think that, yeah, on one head, nothing has changed because, yeah, the bar is still the bar, right? I used to be trek and field runner, almost became a professional, like, trek and field runner, took a different path in life. But my friends went on to the Olympics. and yeah Olympic like running is still the same right you need to train twice a day

you like like push it to the edge and that's not given like to everyone mentally and physically so like that I think has not changed in company building I think what you can do and the output you can generate is just like four or five times but that's the big experiment that we're doing together right it's fun to see okay how quick can we get to all the dentists

in America and build a really good product such that there's 200 hours gone like how quick can we then bring it to another vertical and then help all small businesses? I think that that is the big, interesting experiment that we are going to do over the coming years. If it's so much easier, if everybody can hire a betty, right, just materialize a betty. How does that change? I mean, like, it could actually work out where like small businesses, it's much easier to start one and run one, but then actually, paradoxically, it's much harder to be one.

because you do have, if you think about moats in the AI era in general, we often talk about it with respect to software companies. So it's so easy to go replicate XYZ software, I did it on Replit, or I did it on Lovable, or I did it on Clod. Like you hear this left and right all the time. Much, much harder at the same way to go replicate Dr. Sloop's practice. But one of the things that makes a small business somewhat defensible is actually it is an accumulation of people

that are required to deliver the end product. And it's like if you know who Yogi Berra is, you know, famous Yankees, a baseball player that said all these things that make no sense. And but they... Quoted often, though. Quoted often, right? And one of my very ones, it's so crowded,

nobody goes here anymore. Yeah. It doesn't make sense. But I guess my question is, how do you think small business changes if it's easier to run a small business and start one? Because theoretically, that could erode the market,

of a small business such that it's so crowded, nobody goes here anymore. And like this one moat that exists of just materializing people to deliver a product, it's now so much easier, but therefore it's actually harder. Yeah. I think our current take on that is that this assumes there is a cap on kind of like demand. And if you just look at the dental practice, but the same applies to try to find a good plumber. There's just like twice as much demand that currently can be supplied. And I think this is a great opportunity for, you know, everybody to have great dental care and go twice a year.

And I think the same can be said about primary care doctors. I grew up in the Netherlands. I had a very different primary care experience than like most people in America here. Because it's just really hard to find a good primary care doctor that is in your community, knows you, your family and takes care of you. So I think this we see as an opportunity to kind of like create like now there are 160,000, dentists, hundreds of thousands. of dentists. What if America has half a million dentists or can see kind of like twice as much patients, maybe the same amount of dentists. Plumbing like is the same. I think there's a lot of these like small businesses if they wish they could, you know, bake more pies, but they're constrained

on labor basically. And I think what this unlocks is that people have twice as much time for the craft and that is an exciting future because then all of a sudden, yeah, you're going to see a better world. I think that's that's currently your view, which is super exciting about this technology. I think that's also, to your point, like when we launched and we finally told the world, hey, this is what we've been up to. That was the main piece of like a lot of people talked about that. It's like, wow, an optimistic example of like how this new technology like can be used. Because there's a lot of like what's going to happen to the world. But yeah, nobody really can be against cleaning up busy work for small business owners that should be baking pie.

or polishing nails or cleaning teeth or drilling, depending on who you are, of course. But I think that's so interesting about this. Yeah. So you started the business in 2020, and that was arguably pre-A-I, or pre-what we think of as being this generative AI revolution, which I would call the BC-80 divide of like November of 2023

when ChatGPT launched, right? So if you're, I guess, 2022, sorry. November of 2020 is, when ChatGBT BT launched the public. What is, what are the kind of remaining, so, you know,

then you had reasoning models, you have all of these things that have kind of built on top of the original revolution of, you know, four years ago, call it, like what,

what are the hardest problems to solve? Like, what is it, when we talk about, like, software that does the job of labor, what cannot be done right now? What do you feel like we still need more technical advance to get there?

And sometimes it's like a 90-10 thing, where it's like, you know, the last 10% is really, really hard, but you can't be a feature-complete solution until you've done that. So I'm just kind of curious, like, from a technical lens, what are the things where you would say, okay, and it's not a Lassi-specific question, it's just kind of more of the technology and what it enables writ large. Where does work still need to be done,

where it's just not quite good enough? And then what do you think the curve of that looks like? So it's not, you know, it's like a question of AGI for small business. Yeah, yeah. You know, what do you need? And where are we on that curve if you had an estimate? I think one thing that's interesting is that the models are trained on so much data and they're so large. And yet they actually don't really know how to do any of this work.

Like they don't have the workflows encoded in any way. So, for example, we're working on a product now where we have to like collect all of these like basically SOPs on like and documents about like, how are you? supposed to bill insurance claims to certain payers and all this kind of stuff, which to some extent humans would do the same. But there's also a big amount of just like human knowledge that is encoded in, say, these office managers and they just like know how to do this work. That's weirdly not that accessible on the internet. I think we have a big advantage there because we have all of this like historical data out

of their ERPs that we can look at and kind of infer, you know, some of these workflows. from, but that's something we notice a lot. I think actually when we first sort of using some of the later reasoning models, we kind of assumed like, oh, they'd probably just know how to do this work because, like, why would they not? They're trained on all of this data. But it turns out that they don't know all the intricacies of most of these workflows. I think, yeah, this is less specific to Lassie, but I'm personally kind of excited about

smaller models that, you know, learn faster and can learn online. on less data. I think that will be really cool to see because I, you know, I think over time we'll have intelligence kind of, I think, disseminated everywhere. And, you know, like to the point where like, you know, like the Pixar lamp that has its own personality, like, why not? I feel like I want my intelligence to be to be everywhere. And it feels like we're super far from that. So I'm, you know, curious about that. I don't know if it'll benefit us necessarily that much. but there is something about the models not being able to learn very

very fast and as it relates to us we put a lot of product work into how do you actually get input from you know from people about their preferences and like how they've done the work in a way that we can then

kind of like useful input in a way that we can then use to actually make our agents smarter. And I think from a, you know, like product UX perspective, I think there's a much of interesting stuff to dissolve there as well. Maybe one kind of final question between technical and non-technical is I've been thinking about this a lot, how the world changes when the marginal cost of arguing goes to zero, right? So I was thinking about this because Cigna, who I have for my health insurance, will only send paper checks. I was like, I must have missed the whole like online enrollment. Nope, nope. They don't have one. Why don't?

they have one? Well, they're kind of hoping that, like, you might lose the check, you might not deposit it. It's just like this kind of, like, intentional delay. And it's the same thing for, like, a lot of insurance. It's like, you know, it's like deny, deny, deny. They want to deny. And, like, on the other side, it's like, ooh, you know, pretend that when my house burned down, I had a Picasso in there. Like, both parties are trying to cheat each other. This is not new to the insurance agents or the insurance, like, industry in both sides, right? But now that everybody has this super-powered thing that costs effectively nothing to go argue in perpetuity.

It's like, I can argue with you, and then you can argue with me. But how does that change the business dynamic of things like insurance and payments and collections? I mean, I'm sure you've thought about this a lot. Yeah. Because in many cases, the counterparty that the dentist is dealing with

is the insurance carrier. Yeah. Right. A big part of the cons receivables or revenue comes from insurance companies. Right. Yeah, I think it's quite interesting. because it's pretty well regulated.

So if a dentist does a crown, and you provide them with the correct narrative, Frederick was already talking about it, there is just a document at Cigna that specifies if you give me this X-ray, you give me this narrative, then we will cover that. But for a human, it's quite hard to follow those rules because there are like, take dentistry,

they're like 50, 60, maybe 100 common billing codes. And it's quite a lot to like remember for Mildred or Betty this offer. And in this case, kind of like our agent like has like all the like documentation. It has access to the X-ray that is the relevant like one. It knows how to build like a treatment plan and will then submit that with the insurance company. And insurance company like just needs to follow the rules, right? And then they will like pay it out. The other thing, because indeed we are very deep in this industry.

that is interesting dynamic. It took me a while to understand this, but Cigna also has an incentive to have good doctors in network that are happy with them. Because every year they need to go to Enreason and say, do you want to renew your dental plan with us? And Enrecent is going to look at like, are there good dentists, like in the area that are covered that Alex and Olivia can go to. And if the answer is no, then because there's competition in this market,

they will go to like, you know, another insurance company. So like there is actually an incentive to keep Dr. Slope and Kwan like in network because like otherwise they will not renew the plan like with employers. So like I think that it is a lot more for us it's quite interesting because why I also think these agents are so well equipped to do this work because there's very clear documentation. After Frederick's point and we can talk about this for hours and maybe another time, but like we are also literally digitizing the file cabinet. net. Like, that's, it's really interesting that it's really hard to find these like files.

They're there. And once we have them, we know how kind of like to do this. But all these payments you talked about, a lot of doctors across America still get paid on paper too. So like they get $100,000 in checks like deposited on their desk. And I did this, right? So for Dr. Kwan, I literally knock, knock, knock, open the door. I sit on my bar stool and there's the mailman in this case that gives me a stack. And I kid you not, I open all these envelopes, deposit $100,000 like cash in the bank account of the doctor, and then I have to go to work on all these itemized invoices that are attached to the checks.

And now the interesting thing, so even if we had the models, you couldn't really do this up until a couple of years ago because that was the status quo. So the fall cabinet was literally the fall cabinet. In the file cabinet, like, were the checks and the itemized invoices that you need to do kind of like part of the job to keep the doctor office running.

And then the federal government stepped in, and they said, this has to stop. You cannot do paper checks anymore for much longer. So then they mandated this industry to switch to direct deposits, offer that as an option to a doctor. So if a doctor says, I want to flip the switch, you need to do that. The same is these itemized invoices. You need to create a digital file format like for that.

And we are also writing that like Tillwind. So you will see that like a lot of these small businesses, I think the stats are 70% are still paid on paper. And it's going to this mess of like digitization revolution right now because federal inflection point that's regulatory. So I think that combined with these models being so good makes this super interesting. Because even if you have the models or had the models five years ago,

like, yeah, the payments are still on paper. And we are digitizing that in the meantime, which is also what we talked about earlier. Like it is not one click and then kabum because otherwise the whole country would be on digital payments, right? Like why would kind of like Dr. Sloop not do this? Well, Dr. Sloop needs to also then spend 50 hours figuring out, with Cigna and Delta and all the other insurance companies,

how do I flip the switch? It then comes into the bank account. Do you want the staff for to have access to that bank account? Likely not because the rent payments aren't there, like the other data that you don't want your staff to see. So that's why the status quo is the way it is. And then we built this like massive engine that basically, hey, give us the business info that we need,

the tax ID number and some other nonsense. And then we will go do that conversion to digital payments, which is also a hard product problem to solve. but we have kind of like also product ties like that. So yeah, it's super interesting to think about what other industries like still have that because that's I think even a more interesting mode where it's like, okay, you can apply these models, bring them to Main Street is super hard. And then in addition to that, how do you convert kind of like data that you need

that's currently living in a file cabinet to like digital file format such that you can actually automate the work. Yeah. To your earlier point, actually, about digital file cabinets not being that much more Part of the reason that all of these payments are still on paper is that the staff basically prefers doing the work by hand on paper, right? Because if you just digitize everything, now you have a PDF instead of like a paper in front of you. Like you actually need, you know, some tools to use that, use the PDF. And it's preferable to just kind of, you know, work off of a sheet. And so like there wasn't really much of an incentive to digitize. Now that's obviously different. So obviously there are a lot of dental practices that want or even desperate. really need products like Lassie, but they are distributed and they're all over the country and there are a lot of them to reach. How do you think about reaching them? Like, how do you

bring agents to kind of these mainstream American businesses? Yeah, this is a very different playbook than where currently, I think, the cutting edges is like you have these models good enough and apply them in enterprises and you do like a few steak dinners and then you sign a contract and then you have like 10 million an ARR booked, right? we literally need to go find like thousands, tens of thousands, hundreds of thousands of thousands of small businesses. So it's a super interesting problem solved because we've built as agent that's really good. And what we're doing right now is literally mapping out like where are all these dentists

in this case and after the next small business type in the country, who's the owner? What systems are they on? Like are there any intense signals that we can find, right? Like they're looking for a job because they say it on indeed. and then get in touch with these, like, people with a message that resonates with them. And that cuts to the noise. I think that's, especially with our customer type, is very different, right? Because if you were to sell to, like, me, you like, find me, you know, in clay and you enrich it with some Apollo data.

And then you look me up on LinkedIn. And then, you know, okay, this is the guy that's going to buy my HR system. We can't really do that. Because Dr. Sloop is not in that database. He's often on LinkedIn. in. So like this is a completely different playbook that we're developing here. And I think that's another very compelling kind of like thing that we're figuring out basically what does the go-to-market playbook look like to adopt AI in like many, many businesses. So yeah, that's quite an

exciting opportunity and untapped. Thank you both so much for coming to chat with us today. This was awesome. We are very, very excited for the future of Lassie. Anyone who's listening who might be interested in working with the Lassie team to build something generational here. Check it out at Lassie.AI. And you guys are very actively hiring for what I understand. Oh, yeah. Amazing. Great.

Well, thank you guys again. Thank you so much. Thanks for hosting us. Thanks for listening to this episode of the A16Z podcast. If you like this episode, be sure to like, comment, subscribe, leave us a rating or review, and share it with your friends and family. For more episodes, go to you. YouTube, Apple Podcasts, and Spotify.

Follow us on X and A16Z and subscribe to our substack at A16Z. com. Thanks again for listening and I'll see you in the next episode. As a reminder, the content here is for informational purposes only. Should not be taken as legal business, tax, or investment advice or be used to evaluate any investment or security and is not directed at any investors or potential investors in any A16Z fund. Please note that A16Z and its affiliates may also maintain investment.

and the companies discussed in this podcast. For more details, including a link to our investments, please see A16Z.com forward slash disclosures.