One man built $1.8 Billion company . The Medvi Story

I'm sure most of you have seen the New York Times piece by now about Matthew Gallagher and Medvi. Guy launches a telehealth company from his living room in LA with $20K and a bunch of AI tools, hits $401 million in his first year selling GLP-1 weight loss drugs, and is now tracking toward $1.8 billion in 2026 with literally two full time employees (him and his brother). Sam Altman apparently won a bet with his tech CEO friends over when the first one person billion dollar company would show up.

Three people in my circle have already sent me the article with some version of AI can do anything if prompted properly . I spent an hour on the phone with him last night trying to add some nuance, and I figured it was worth writing out my full thinking here because I imagine a lot of people in this sub are processing the same thing.

Gallagher used ChatGPT, Claude, and Grok for code and copy. Custom AI agents for connecting his systems and monitoring performance. One person doing the work of what would have been a 15 person team across engineering, design, marketing, and support a few years ago.

But Gallagher didn't actually build a telehealth company from scratch. He built an extremely efficient customer acquisition and marketing front end on top of two existing infrastructure companies, CareValidate and OpenLoop Health, who handle the actually hard and expensive parts: licensed physicians, prescription processing, pharmacy fulfillment, shipping logistics, and all the regulatory compliance. OpenLoop is a real company out of Iowa with real employees doing all the regulated work. As someone on X pointed out, "solo doesn't mean what people think it means in 2026" when your entire operation runs on top of another company's infrastructure.

He also picked maybe the most explosive consumer market of the decade. GLP-1 weight loss drugs have near infinite demand, customers are desperate to buy, and the supply of affordable options through traditional channels is constrained.

This story is probably more about how powerful the GLP-1 market is than about how powerful AI is. Gallagher's marketing instincts and his timing were arguably more important than his AI toolkit.

And there are real cracks showing with this fast growth. The FDA sent warning letters to dozens of telehealth companies including Medvi in March 2026 over concerns about the compounded GLP-1 drugs being sold.

His own AI customer service chatbot was fabricating drug prices and hallucinating product lines that didn't exist early on. He honored the fake prices, which is to his credit, but that's still a concerning failure mode when you're dealing with pharmaceuticals.

None of this takes away from the fact that the result is extraordinary. $401M in year one with two people is genuinely unheard of. But the interesting question isn't whether this is impressive (it obviously is), it's what lesson is actually transferable to the rest of us who aren't sitting on top of a once in a decade consumer demand wave.

This is something I've been thinking about a lot because I run a B2B company with a team of 3 people, and we've been living the "AI lean team" reality for about a year now, just at a much more modest scale. We use AI across nearly everything: content, lead research, data enrichment, parts of customer onboarding. For outbound sales specifically so that we can focus on product, we consolidated from this sprawling mess of separate tools for data, email sequencing, phone, LinkedIn automation, and CRM into a single platform Fuse ai to handle most of the pipeline generation work that used to require at least 2 dedicated SDRs. Three people doing the work of what would have been 8 to 10 two years ago and for us the leverage is just extraordinary on everyday basis

Now I believe the leverage isn't in replacing your whole team with AI. It's in figuring out which parts of your operation are essentially distribution and execution work (marketing, outreach, content production, data processing, basic customer service) and compressing those with AI, while investing more in the things that require actual human judgment. In our case, AI handles maybe 70% of the total work volume but the 30% our 3 humans do is all high judgment stuff: choosing which markets to go after, building real relationships with key accounts, making product decisions, and critically, catching the mistakes AI makes before they reach customers.

One thing I'm taking from the Medvi story is that the businesses best positioned to ride this wave are the ones that can plug into existing infrastructure for the hard parts. Gallagher didn't solve telehealth compliance or pharmacy fulfillment or physician licensing. He found companies that had already solved those problems and built a better front end on top of them. If you're thinking about going lean with AI, the first question isn't "what AI tools should I use" but "what infrastructure already exists in my space that I can build a distribution layer for?"

I'd love to hear how others are thinking about this. Are you running a meaningfully smaller team because of AI, or is it more that AI is making your existing team faster? For anyone genuinely considering going solo because of stories like this, what's your plan .