Alex Rampell

1. Key Themes

Software's Evolution: From Filing Cabinet to Labor Replacement

Alex Rampell articulates a sweeping thesis about the three eras of software: first, digitizing filing cabinets (Sabre, PeopleSoft, QuickBooks); second, bundling financial transactions to expand addressable markets (Toast, Stripe); and third — and most transformative — software that actually does the work. The key insight is that the first two eras barely improved productivity.

"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... did PeopleSoft 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." [00:08:28]

"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 in a way that 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." [00:11:36]

The Labor Shortage Is the Real Market Opportunity — Not AI Displacing Jobs

Both founders and Alex Rampell converge on a counterintuitive point: AI in SMBs isn't taking jobs — it's filling a vacuum that already exists because workers can't be found or retained. Dr. Sloop retired not because AI displaced him, but because his key administrative employee left and he couldn't replace her.

"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." [00:00:53]

"There are about 160,000 dental practices in the U.S. alone, and they spend roughly $200,000 a year on administrative costs. And they can't find people. So it is the doctor themselves with their Harvard degree that sits there till midnight." [00:14:32]

AI Is Overhyped in Silicon Valley, Underhyped on Main Street

There's a profound asymmetry in where AI awareness and adoption currently sits. The businesses that need it most — dentists, gastroenterologists, plumbers, primary care doctors — are the hardest to reach and the least plugged into the AI hype cycle.

"AI is overhyped in Silicon Valley, but underhyped in Iowa. And there are a lot of people in Iowa." [00:00:00]

"When we finally told the world, hey, this is what we've been up to, that was the main piece — a lot of people talked about it. It's like, wow, an optimistic example of how this new technology can be used, because there's a lot of 'what's going to happen to the world.' But nobody really can be against cleaning up busy work for small business owners that should be baking pies or polishing nails or cleaning teeth." [00:44:08]

The "Do the Job First, Then Automate" Product Strategy

Lassie's founding approach — literally sitting in dental offices doing the administrative work by hand before writing a single line of production code — produced a product that actually works autonomously, a rare thing in AI software. This is a distinct and replicable playbook.

"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." [00:17:48]

"Initially, we were actually the humans in the loop. 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." [00:00:13]

No Software Incumbent = Extraordinary Defensibility for AI-Native Entrants

The dental and broader SMB healthcare administration space has no meaningful software incumbent. The real competitor is human labor — "Betty" — which means there's no Workday or Salesforce to copy the feature and leverage existing distribution. This is one of the most defensible moats in the AI era.

"The incumbent was named Betty and she quit two weeks ago. That's the incumbent." [00:27:24]

"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." [00:00:30]

The Mandatory Digitization Tailwind Nobody Is Talking About

The federal government has mandated that insurance companies offer electronic payments and digital remittance to medical providers — moving an industry where 70% of payments are still paper checks into digital rails. This regulatory inflection point is independent of AI and massively amplifies Lassie's opportunity.

"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. The same as these itemized invoices, you need to create a digital file format for that." [00:52:57]

"A lot of these small businesses — the stats are 70% are still paid on paper. And it's going to be this massive digitization revolution right now because of a federal inflection point that's regulatory." [00:53:27]

Pricing Against Labor Budgets, Not Software Budgets

Lassie prices its product against the labor line item on a P&L, not against software subscription comparables. This dramatically expands willingness to pay and reframes the competitive set.

"We are already charging five figures for this first agent that only does 30 hours of labor a month. And there's 200 hours of labor to be done for Dr. Sloop. It comes out of the P&L on the labor budget." [00:15:57]

Building a Consumer-Grade Onboarding for B2B AI Agents

The hardest unsolved problem in bringing AI to SMBs isn't the AI — it's the onboarding. Lassie solved this by building a near-self-serve setup flow modeled on Robinhood and Coinbase's consumer KYC/account-linking flows, abstracting all complexity under the hood.

"The doctor in Iowa says, yes, I want this. They then go to almost like a Stripe-like checkout or a Robinhood-like onboarding flow where they hook up the bank account of the practice, they link the system of record, they link all the insurance portals that claims come in from. It confirms business information. And then under the hood it configures things." [00:31:59]

"One of the many things that Robinhood 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." [00:32:28]

The Marginal Cost of Arguing Going to Zero Reshapes Insurance Dynamics

As AI agents can now argue insurance claims in perpetuity at near-zero cost, the entire economics of claim denial as a strategy begins to break down. Simultaneously, properly documented claims get paid more reliably.

"I've been thinking about this a lot, how the world changes when the marginal cost of arguing goes to zero... Cigna will only send paper checks. Why don't they have one? Well, they're kind of hoping that you might lose the check. You might not deposit it. It's just like this intentional delay." [00:48:58]

"Our agent has all the documentation. It has access to the x-ray that is the relevant one. It knows how to build a treatment plan. It will then submit that with the insurance company and the insurance company just needs to follow the rules." [00:50:37]

AI-Native Hiring: The "AI-Builtness" Filter

Lassie deliberately screens for what Steijn calls "AI-builtness" — whether a candidate fundamentally believes and operates as if AI will change every function. Combined with unchanged standards for raw talent, this creates a team that can move 4-5x faster than a traditional team.

"We see a pretty clear division — do you believe that the way you code will change completely as a result of building a company in this era, building out a finance department? We interview specifically for that, because we want to build the 2026 version of a big organization, where we ship twice as much as others, we move four times as fast." [00:40:12]

2. Contrarian Perspectives

SMB AI Adoption Has Almost No Skepticism — The Problem Is Access, Not Receptivity

The conventional wisdom is that small business owners are slow, tech-resistant, and hard to convert to AI tools. The Lassie founders found the exact opposite: SMB owners are desperate for solutions and adopt quickly once they see it working. The barrier is simply reaching them.

"A lot of people are also surprised. But isn't there a lot of skepticism? 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. We come by and we say, hey, we actually have built this agent that can provide you already with tens of hours of labor. They then adopt it very quick." [00:15:00]

Software Didn't Actually Make the World More Efficient

Contrary to the entire technology industry's self-image, Rampell argues that the first 50 years of enterprise software produced essentially zero productivity gains — the same headcount did the same work, just with a computer instead of a filing cabinet.

"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." [00:08:56]

The "Boring Pipe" Strategy Beats the "Smart Feature" Strategy for Startups

Against the instinct to build exciting, differentiated features, Rampell argues that the right strategy is almost always to first own the boring infrastructure and customer relationship, and only then layer on the lucrative product. He uses his own regret about TrialPay not becoming Stripe as a cautionary example.

"The thing that a lot of startups should do is they should do the boring thing. They should build the raw pipes, the raw — just own the customer. And then you get to build the fun feature on top... I realized what I should build is this thing called Stripe. And this was not revisionist history because Stripe had five people." [00:23:11]

Fintech's Market Expansion Was Modest Compared to What's Coming

Fintech is widely celebrated as the big market-expanding force of the last decade. Rampell argues it was actually a rounding error compared to AI doing labor — and the dental market alone illustrates this numerically (160,000 practices × $200,000 in labor = $32 billion addressable in a single specialty).

"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 labor... it's like, yeah, fintech made it a little bit bigger. But now we can do work. And we can charge for work." [00:11:15]

Making Small Businesses Easier to Run May Paradoxically Make It Harder to Own One

As AI removes the labor constraint that previously protected small business operators from competition, the moat of "you need people to deliver this service" evaporates. Rampell raises the Yogi Berra paradox: if everyone can run a small business, will margins compress so badly that no one can?

"If everybody can hire a Betty — just materialize a Betty — how does that change... theoretically that could erode the margin of a small business such that it's so crowded, nobody goes here anymore." [00:41:40]

Steijn's counter: demand is uncapped. There are twice as many patients who need dental care as can currently be served — AI expands supply to meet latent demand rather than increasing competition for fixed demand. [00:43:13]

3. Companies Identified

Lassie

AI agent platform automating administrative work (insurance billing, patient payments, claims reconciliation) for healthcare practices, starting with dental. Currently serving hundreds of practices at 98% automation rates, charging five figures for its first agent module covering 30 hours of labor per month. Reached most growth via word-of-mouth among dental professionals.

"We are already charging five figures for this first agent that only does 30 hours of labor a month. And there's 200 hours of labor to be done. They see us as someone they break in to actually run the practice for them." [00:15:57]

Robinhood

Consumer brokerage that pioneered self-serve financial account onboarding, KYC, and bank linking — cited as the template for Lassie's SMB onboarding flow.

"One of the many things that Robinhood 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." [00:32:28]

Superhuman

Email client known for its high-quality onboarding, strict ICP discipline, and obsessive focus on delivering core product value quickly — Frédéric Renken worked there and directly applied its playbook to Lassie.

"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." [00:36:20]

Toast

Restaurant software and payments platform cited as the canonical example of how bundling fintech with vertical software unlocks markets that couldn't support pure software pricing.

"Toast could have existed in 1985. But if you bundle in payment processing, you're effectively charging $100,000 for that... fintech made the market much, much bigger for software because of this bundling effect." [00:10:19]

Stripe

Payments infrastructure company cited as the model for "own the boring pipe first" strategy and also referenced as the design inspiration for Lassie's checkout onboarding flow.

"We should do boring payment processing, which is a commodity business. Because if we do that, then we own the customer... the doctor in Iowa goes to almost like a Stripe-like checkout." [00:23:41]

Sabre Systems

Joint IBM/American Airlines project (1960s) that Rampell identifies as the true time-zero origin of enterprise software — digitizing airline reservations from paper filing cabinets.

"Pick the time equals zero moment for that with Sabre Systems. Because airlines would just keep reservations in filing cabinets. Sabre Systems was a joint project between IBM and American Airlines. That's why Sabre is spelled with two A's." [00:07:31]

Coinbase

Crypto exchange cited by Steijn as the consumer onboarding benchmark Lassie aims to match in simplicity for SMB setup.

"The onboarding needs to be as simple as onboarding on Coinbase or Stripe." [00:34:39]

Workday

Enterprise HR software cited as the archetypal incumbent that an AI-first startup must either outmaneuver or avoid competing with directly.

"If 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 — it's such a great idea that this thing called Workday might copy. And they own all the customers." [00:24:08]

Cigna

Health insurer cited as a real-world example of intentional payment friction (paper checks only) that AI agents can now systematically overcome, and also as a stakeholder with aligned incentives to keep quality doctors in-network.

"Cigna will only send paper checks. Why don't they have one? Well, they're kind of hoping that you might lose the check. You might not deposit it. It's just like this intentional delay." [00:48:58]

TiVo

Digital video recorder company cited as the canonical "TiVo Problem" — a great innovation trapped in a terrible competitive position because distribution is owned by incumbents.

"TiVo famously, TiVo and Replay TV both invented the digital video recorder... but a terrible company because you really have very few outcomes that are good." [00:22:16]

PeopleSoft / Workday / NetSuite / QuickBooks / LexisNexis

Cited collectively as examples of the first wave of enterprise software — digitizing filing cabinets without meaningfully improving productivity.

"There are HR filing cabinets and that became something like PeopleSoft. There are legal filing cabinets and that became all of these LexisNexis products. There are accounting filing cabinets and that became QuickBooks and that became NetSuite." [00:07:31]

4. People Identified

Steijn Pelle

Co-founder and CEO of Lassie. Former Robinhood growth/referral program. Originally from Amsterdam. Track and field athlete (near-professional level). Identified the dental administration problem by personally doing the work in Dr. Kwon's office — opening mail, depositing checks, reconciling invoices — before building any product.

"I never forgot what I saw. A small business owner that is the number one rated doctor on Yelp, spending 200 hours a month on paperwork and busy work." [00:00:13]

Frédéric Renken

Co-founder of Lassie. Former product lead at Superhuman. From Hamburg. Responsible for the technical architecture — built the context layer, tool integrations, and agent intelligence stack starting in 2020, positioning Lassie to leverage improving models as a tailwind.

"We started building the context layer and building the tools... as the models got better, we had this huge tailwind because we had all this context already built, all the tools already built. And we could, as the models got better, just replace our intelligence and the product would just get smarter over time." [00:06:35]

General Partner at a16z. Former founder of TrialPay (sold to Visa). Credited with first articulating the thesis that software would do the job of labor. Originator of the "TiVo Problem" and the "startup vs. incumbent distribution race" frameworks.

"The battle between every startup and incumbent comes down to whether the startup gets the distribution before the incumbent gets the innovation." [00:00:30]

Dr. Kwon

Lassie's origin customer. Number one rated dentist on Yelp in his area. Working 200 hours a month on paperwork while running a thriving practice. Gave Steijn and Frédéric full access to his business operations.

"There was a great quote from Dr. Kwon about you guys, which is that Lassie isn't replacing humans, but 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." [00:00:46]

Dr. Ronald Sloop

Alex Rampell's family dentist in Florida. Retired not due to age but because his key administrative employee left and he couldn't manage the practice alone — cited as the definitive human example of the labor shortage problem Lassie solves.

"Part of why he retired was he lost his key woman that did the books and everything else. He's like, I can't deal with this anymore. I quit... He said that if this had been around, he wouldn't have retired." [00:12:23]

Dr. Shah

Gastroenterologist in Scranton, Pennsylvania. Second practice Lassie founders worked in hands-on to validate that Dr. Kwon's administrative chaos was systemic, not anomalous.

"I also worked for a gastroenterologist in Scranton, Pennsylvania. And we saw the same there. I'm like, wait, you're doing this all by hand?" [00:03:49]

5. Operating Insights

Set Automation Thresholds Before Launch, Not After

Frédéric articulates a disciplined product release standard: don't sell a product until you've reached ~95%+ automation on that specific workflow. Don't wait for 100%, but also don't ship something that still requires significant human involvement. This prevents getting "stuck" with customers you've over-promised and can't serve autonomously.

"We want to get to a sufficient level of automation across any product before we sell it. For us, that's like 95 plus. But not necessarily 100. I don't think we're going to wait until we get to 100 with one product and then do the next one. As soon as we can take over a job, we take it over and then we move over to the next one." [00:18:47]

Strict ICP Enforcement Is More Critical for AI Agents Than Traditional SaaS

In traditional SaaS, a bad-fit customer is a churn risk. In autonomous AI agent products, a bad-fit customer can trap you in an unresolvable commitment to do work you can't actually automate yet. The cost of onboarding the wrong customer is disproportionately high.

"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 we claimed we're going to automate a bunch of labor for them. We can't do it. Are we going to offboard them? It's particularly painful, maybe more painful than in a kind of old world product." [00:36:20]

Build Onboarding as a Scripted Movie, Not a Feature Set

Frédéric describes treating the onboarding journey as a narrative with hard checkpoints and defined timeframes — actively measuring whether each "scene" is hit on schedule, rather than letting customers self-navigate to value.

"Our onboarding is a little bit like a story or a playbook or like a movie. We have set points and checkpoints that we want to reach in certain timeframes and we make sure it happens every time. And we measure that of course." [00:36:49]

Staff Feedback Loops as a Systematic Model Improvement Engine

Lassie has turned its customer base's human staff — who handle the residual 2% of cases the agent can't yet process — into a structured data flywheel for improving agent accuracy. This is a moat that deepens with every customer added.

"We have thousands of staffers that are basically giving us input on how to make that appeal that the agent currently cannot do." [00:21:04]

Interview Explicitly for "AI-Builtness," Not Just AI Familiarity

Beyond general AI enthusiasm, Lassie screens for candidates who fundamentally believe their specific function (engineering, finance, sales) will be structurally transformed by AI — and who operate that way daily. This filters for people who will build and operate the 2026 version of a company, not a 2019 company with AI bolted on.

"We see a pretty clear, a clear division — do you believe that the way you code will change completely as a result of building a company in this era, building out a finance department? We interview specifically for that, because we want to build the 2026 version of a big organization, where we ship twice as much as others, we move four times as fast." [00:39:46]

6. Overlooked Insights

The SMB Go-to-Market Playbook for AI Is an Entirely Unsolved, Massive Prize

Steijn briefly mentions — almost as an aside — that Lassie is building something that has never existed: a scalable system for finding, reaching, and converting non-LinkedIn, non-database SMB owners at scale using intent signals (like job postings on Indeed). This is not a solved problem. Every AI company targeting SMBs faces this exact bottleneck, and whoever cracks the playbook owns a distribution advantage across every vertical simultaneously.

"We are literally mapping out where are all these dentists in the U.S., then after that the next small business type — who's the owner, what systems are they on, are there any intent signals that we can find, like they're looking for a job because they say that on Indeed. Dr. Sloop is not in that database. He's often not on LinkedIn. So this is a completely different playbook that we're developing here." [00:56:24]

This is arguably as valuable as the Lassie product itself — a replicable demand-gen engine for the entire "AI for Main Street" category.

The Proprietary Workflow Ontology Is the Real Moat — Not the AI

Frédéric makes a quiet but devastating observation: even the best frontier models don't know how to do medical billing. The tacit knowledge of how to actually process an insurance claim is not on the internet. Lassie has spent years extracting this knowledge from human staff, encoding it into structured workflows, and building a proprietary ontology that maps across incompatible systems. This data asset — not the model — is what makes Lassie impossible to replicate quickly.

"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... there's 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. We have a big advantage there because we have all of this historical data out of their ERPs that we can look at and kind of infer some of these workflows from." [00:00:13]