Feb 12, 2026 · 49m · product-market-fit
How I grew my AI startup to $3M ARR in 3 months. | Chaz Englander, Founder of ModelML · PMF Show
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of The Product Market Fit Show, serial entrepreneur Chaz Englander shares actionable frameworks and playbooks behind building and selling three multi-million dollar startups: Fat Llama, Fancy, and ModelML. He breaks down scrappy MVP testing, enterprise AI workflow scaling, tactical fundraising, and the continuous effort required to maintain product-market fit.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Pablo holds 18.4% of the talking time here. How this is scored →
Pablo as informed peer 3.8 Guest teaching 4.3 Guest disagreement 1.2 Pablo pushing back 1.2
speaking balance: gold is Pablo, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 10:23 Rejection of the no-competition myth
Chaz strongly dismisses the conventional first-time founder anxiety over competitors, asserting bluntly that if nobody else is building in your space, you probably lack a real market.
Hardest push from Pablo ▶ 29:31 Challenging guest's edge in personal asset managementPablo directly challenges Chaz on whether starting a personal investment office was naive given how hard it is to generate alpha without an established market edge.
Biggest teaching moment ▶ 36:27 Explaining agentic cognitive flows in enterprise workflowsChaz educates the host on how Model ML maps agentic systems directly to human cognitive flow across information gathering, creation of 200-page decks, and automated verification.
Pablo holds their own ▶ 39:43 Citing Leya case study to validate embedded design partnershipPablo displays sharp industry knowledge by immediately matching Chaz's on-site design partner thesis to the playbook used by Leya to scale to a unicorn valuation.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Pablo as informed peer | Guest teaching | Guest disagreement | Pablo pushing back | Why |
|---|---|---|---|---|---|---|
| The Origin of Fat Llama and Scrappy Early Fundraising | 4 | 3 | 1 | 1 | Pablo shares his own experience getting into 500 Startups from Ottawa to relate to Chaz's London to YC journey. Chaz explains how he cold-pitched LinkedIn contacts in Canary Wharf for small angel checks. | |
| First-Time Founder Fundraising Playbook & Narrative Framing | 4 | 5 | 2 | 1 | Chaz lays down actionable rules for first-time fundraising, emphasizing that founders must never claim they are raising from zero. Pablo contributes his own accidental momentum strategy during fundraising. | |
| Perseverance, Market Validation, and Overcoming Fear of Competition | 4 | 5 | 2 | 1 | Chaz reframes the fear of competition, arguing that having no competitors usually means having no viable market. Pablo validates the distinction between blind perseverance and working through inevitable startup obstacles. | |
| Scaling Fat Llama, Unit Economics, and Dynamic Product-Market Fit | 3 | 4 | 1 | 1 | Chaz details the three-year grind to fix unit economics at Fat Llama, distinguishing category expansion from geographic expansion. He introduces the concept that product-market fit is fluid and can be lost quickly in AI. | |
| The Collapsed SPAC Deal and Successful Exit of Fat Llama | 3 | 3 | 1 | 1 | Chaz recounts the emotional weight of a 9-month SPAC merger failing just before Christmas, followed immediately by an acquisition offer from a Swedish group. | |
| Launching Fancy and Extreme Scrappy MVP Execution | 3 | 5 | 1 | 1 | Chaz describes the extreme scrappiness of launching Fancy in Newcastle with Twilio texts and corner-shop runs. Pablo probes whether they had existing merchant partnerships. | |
| Testing Core Demand and Modern MVP Development in the AI Era | 4 | 4 | 1 | 1 | Pablo and Chaz discuss how modern tools like vibe coding collapse MVP build times from months to a single day, eliminating the technical excuse for not launching. | |
| Scaling Fancy, Financial Pressure, and the GoPuff Acquisition | 3 | 4 | 1 | 1 | Chaz outlines the rapid scaling and intense financial stress of Fancy expanding across 15 cities while burning capital right before being acquired by GoPuff. | |
| Origin of Model ML: Transitioning from Investing to Software | 5 | 4 | 1 | 3 | Pablo pushes back with skepticism on whether setting up a family office and investing their own capital without an established edge was naive. Chaz concedes they broke even but built internal automation software out of necessity. | |
| Model ML Product Breakdown: Enterprise AI Workflow Automation | 3 | 5 | 1 | 1 | Chaz breaks down Model ML's product architecture, detailing three agentic systems mimicking human cognitive flow to automate complex finance decks and verification. | |
| The Design Partner Playbook and Co-Locating with Enterprise Users | 5 | 4 | 1 | 1 | Chaz emphasizes that founders must physically sit inside their design partner's office. Pablo cites Leya/Legora as an exact parallel case study in legal AI. | |
| Measuring True Engagement, Hitting Escape Velocity & $3M ARR | 4 | 5 | 1 | 1 | Chaz outlines reaching escape velocity, growing from $5k to $100k MRR in 3 months, and explains why enterprise sales differentiation relies entirely on credibility and execution speed. |
Statements from this episode (23)
Englander: Fat Llama raised £1.1M across two rounds before Y Combinator
“I'd say that one of the things that we don't talk about enough was like, you know, between starting and getting on YC, we raised the million pounds and that was like straight scrappiness. I remember we actually raised two rounds, a 100,000 pound round, and the…”
Englander: 90% of Fat Llama's early capital came from secondary intros
“Probably only five, 10% of that came direct from those LinkedIn messages I was sending. The other 90% was indirectly, almost entirely through that, right? Because, you know, at the end of the day, if you're meeting one person, you're not meeting just them. You…”
Englander: Founders must always frame funding rounds as nearly completed
“It sounds obvious, but, like, I would advise, like, not to do that ever, right? In other words, like, no matter what, if you are raising, the round's pretty much already done. That's the narrative.”
Englander: Founders should set round targets to first check plus 20%
“That's what people should do is whatever their first check is in the round. Just say your total round is that plus like 20%.”
Englander: Having zero competition usually means there is no customer demand
“If you have no competition, and you've like never had any competition, it's like you still have no competition. You probably don't have anything that people want.”
Englander: Out-executing competitors is easier than finding a working concept
“It's way easier to execute better than it is to find a concept that works.”
Englander: Fat Llama fixed unit economics a year after $12M Series A
“We got on YC, raised a twelve million dollar Series A. Pretty quickly after that, but it took us a good year or so after our series A for the unique economics to add up”
Englander: Fat Llama was doing just under $1M monthly GMV at acquisition
“I think when we sold it was about, I might be speaking out of turn here, just under a million a month in GMB, like there or thereabouts.”
Englander: Product-market fit in AI is not static and can disappear overnight
“Product market fit is not necessarily static as well. I think particularly now in the AI world, you might have a PMF today, but you don't necessarily have it when you wake up.”
Englander: Fat Llama spent over $500K on a collapsed SPAC listing
“We spent over half a million in legal fees, and, you know, nine months of work, you know, me in the middle of the night, all the time with my laptop on my knees, two, three weeks before Christmas, the whole thing just fell out of bed.”
Englander: Fancy's founders personally completed its first 1,400 deliveries
“The team is in the founding team did the first like 1400 deliveries.”
Englander: Fancy reached thousands of daily orders pre-COVID without marketing
“So we quickly got to like hundreds and then thousands of orders a day with no marketing, right? So it was purely, it was a referral-based Thing and sticky and just so different to the first company.”
Englander: Vibe coding is now ready for production software development
“I mean, a year ago, vibe coding was great for prototyping. I mean, now I think it's, like, great for production. So, in terms of MVPs, like, the concept of saying I don't have technical resources to build an MVP is just, like, not true.”
Englander: Fancy expanded from 0 to 15 UK cities in six months
“We went from zero cities to I think 15 cities in about six months.”
Englander: Family office broke even and made no net money
“If you're asking whether we made money doing it, no. On balance, no. But we didn't lose money. I, but we did make some money on some things. And then we did do some great investments and a couple of things. But generally speaking, like we weren't making any mo…”
Englander: ModelML rejects prospective customers spending less than $250,000 annually
“I mean, all the way to the extent now we, partly due to the demand, partly due to the way that our business model is evolving, you know, we don't really look at customers that are going to be spending any less than maybe a quarter of a million dollars a year w…”
Englander: B2B founders will fail unless they work from design partners' offices
“If you've got a design partner that you're speaking to once a week, you're almost certainly gonna fail, right? If you've got a design partner that you are working from their office, and ideally sat next to them, you're almost certainly going to succeed.”
Englander: Shipping and iteration speed is a startup's only competitive advantage
“Really your only tools, the speed at which you ship product and the speed at which you iterate, you know, that's really it. That's like kind of the only competitive advantage to be honest.”
Englander: Real product feedback is unprompted solo usage, not demos
“Real feedback is not when you are showing the user the product or what they can do. Real feedback is where they are using it, and ideally by themselves, right?”
Englander: ModelML worked a 997 schedule for their entire first year
“First year, we were nine, nine, seven. Whole year. Every waking hour, frankly.”
Englander: ModelML grew from $5K to $100K monthly revenue in 3 months
“We went from about five K in monthly revenue To about a hundred K in three months. And then we kind of did that again in like the next three months.”
Englander: ModelML raised a $75M round despite spending almost nothing
“We've just raised big, random, We're at 75. We raised 15 before, and we basically spend no money, right?”
Englander: Foundation models are not an existential risk to vertical AI startups
“Our risk is, as I mentioned just then, is like, we may not have product market fit tomorrow morning. And that's not based on, like, OpenAI or Anthropical. It's just based on the speed that Verticalize application is shipping great products. So really, our only…”