In September 2024, Matthew Gallagher launched a telehealth company from his house in Los Angeles with $20,000 and a stack of AI tools. By the end of 2025 it had done $401 million in revenue, served 250,000 customers, and generated $65 million in profit — a 16.2% net margin that publicly traded competitors with thousands of employees can only dream about.
The New York Times called it proof that Sam Altman's prediction had come true.
Sam Altman, for reference, had said in 2024: "In my little group chat with my tech CEO friends, there's this betting pool for the first year that there is a one-person billion-dollar company."
Medvi — with two employees, both brothers — is as close as we've got.
Here's the part I can't stop thinking about.
The build
Gallagher didn't invent the idea. Telehealth for GLP-1 weight loss drugs — Hims and Ro had been doing it for years with thousands of employees and hundreds of millions in funding. His bet was structural, not inventive.
He used AI for every layer he could control directly. ChatGPT, Claude, and Grok to write code and website copy. Midjourney and Runway for ad creative. ElevenLabs for customer communications. Custom AI agents to connect his internal systems. He even cloned his own voice to handle personal scheduling — every recovered hour went back into the business.
For the parts he couldn't control — the regulated, complex, slow-moving infrastructure — he didn't try to build them. He found CareValidate and OpenLoop Health, existing telehealth-in-a-box platforms that provide the doctor network, pharmacy connections, compliance, and shipping. He plugged in, owned the customer acquisition layer, and let someone else run the hard part.
The numbers that resulted are almost absurd. $3 million in daily revenue. 250,000 customers. $65 million in profit. A 16.2% net margin versus Hims & Hers' 5.5% with 2,400 employees. Yahoo!
And he answered over 1,000 customer service calls personally, because his AI chatbot was routing unhappy customers directly to his cellphone.
The real story under the story
Six weeks before the NYT profile ran, the FDA had sent Medvi a formal warning letter for misbranding violations. His clinical partner OpenLoop had suffered a data breach in January 2026 exposing 1.6 million patient records. A class action lawsuit had been filed months earlier. And a Futurism investigation had documented over 800 fake doctor accounts on Facebook running Medvi ads — AI-generated personas with fabricated medical credentials.
The BBB gives Medvi an F rating. Reddit complaints reveal patterns around refund policies, billing practices, and customer service that Trustpilot's 4.5 stars don't capture. The gap between positive Trustpilot reviews and critical Reddit posts suggests experiences vary significantly. PeterMD
The New York Times verified the revenue. They didn't verify the distribution playbook that generated it.
The expansion that raises more questions
Here's what's interesting about where Gallagher went next. Rather than reinforcing the foundation, he's reinvested profits into rapid expansion: men's health launched February 2026 and signed 50,000 customers in its first month. Meal delivery launched in March. Women's health, hair growth, and skincare are queued. Yahoo!
That's five new verticals in six months, from a company with two employees and unresolved FDA, litigation, and data breach exposure in its original business.
It's a pattern worth naming: when you build on a market tailwind rather than a structural moat, the instinct is to move fast and expand before the window closes. Because you know, even if you don't say it out loud, that the window might close.
The question nobody's asking
The internet is full of content right now about how AI makes it possible to build a $20K MRR app in six weeks. Or an agent that automates your marketing. Or a solo business that hits $1M ARR.
Medvi is the ultimate version of that story. And it might be the perfect illustration of why that framing is incomplete.
Getting to revenue is no longer the hard part. Any founder with a decent idea, genuine AI execution, and a willingness to move fast can get there. The market has proven that. Medvi proved it at an extreme scale.
The question that nobody's asking is: what happens in year three?
Every business has a lifecycle. The ones that create real value — that weather competitors, regulatory shifts, market changes, and the inevitable moment when the tailwind stops — are built on something more durable than a growth channel and a market surge.
Medvi's Right to Win was real: Gallagher's marketing expertise, genuine AI fluency, and the insight to rent infrastructure instead of building it. That's a legitimate structural advantage. But the growth engine that got him to $401M ran on fake doctor personas, regulatory grey zones, and a GLP-1 market that was exploding regardless of who was selling. The FDA is catching up. Compounding pharmacy regulations are tightening. Novo Nordisk's new Wegovy subscription program is narrowing the price gap that made Medvi's model so attractive. And he's expanding into five new verticals simultaneously with two employees and active legal exposure.
That's not a business being built to last. That's a revenue spike being extended.
"Getting to revenue is no longer the hard part. The question nobody's asking is what happens in year three." - RogueFounder
RF Take: Medvi is an most important case study — not because of the $401M, but because of what it reveals about the current moment in AI-native building. We're in the first chapter of a shift where getting to revenue genuinely is easy. AI has compressed the startup playbook so aggressively that a smart founder with $20K can generate nine figures. That's real. What hasn't changed is what it takes to build a business that lasts — real structural moats, genuine customer trust, regulatory resilience, a reason to exist beyond a market tailwind. The RogueFounder question isn't just how do you get to revenue with AI. It's how do you build something with intrinsic value — a business that would survive if the tailwind stopped tomorrow. Medvi got to $401M. Whether it exists in three years is the more interesting question.