It was Christmas morning in Kenya. The elephants were scheduled to arrive by 9 a.m., and Emilian Popa, the co-founder and CEO of Expand Health AI, was already coding.
“My wife and kids were kite surfing at the beach while I was building,” he says. “We’d do safaris during the day, and I’d be at my laptop before the elephants showed up in the morning.”
He was not coding for fun. He was finally building the platform his company had spent close to a million dollars and two years trying to build with a full development team.
Six weeks later, it was live. A fully functional, compliance-ready AI platform serving real patients across three continents. And it was built alone over a holiday by a founder who had taught himself to code months earlier.
When the vision is clear, but the build keeps breaking
Have you ever known exactly what you needed to build but had no way to build it? That is where Emilian had been stuck for two years.
He is a co-founder and CEO of Expand Health AI, operating functional medicine clinics in Cape Town, Bucharest, and Miami. Functional medicine means deep, personalised care: bloodwork, hormone panels, gut tests, heavy metal screening, DNA reports, wearable data, all of it analysed and connected into a treatment protocol for each patient.
The problem was the time it took.
Before the platform existed, a single patient visit consumed roughly seven hours of a clinician’s time. The patient arrived with up to thirty pages of lab results. The clinician read every page manually, cross-referencing biomarkers against functional medicine ranges, connecting patterns across systems, writing a protocol from scratch, drafting a plain-language summary, formatting it, and sending it.
“Most of this was done in Word documents, spreadsheets, and sticky notes,” Emilian explains. “There was no system connecting the labs to the protocol to the patient communication.”
A million dollars and nothing to show for it
Emilian knew the solution: an AI platform that could read the labs, extract every biomarker, flag what was out of range, and generate a draft protocol before the clinician entered the room. He had the clinical expertise, the patient volume, and the practitioner network. What he could not do was build it.
The company spent close to a million dollars attempting to solve this with a traditional development team. The models were not ready, and the retrieval architecture was not there. Every iteration moved slowly and expensively.
“We failed,” Emilian says simply. “The platform we needed simply couldn’t be built the way we were trying to build it.”
Two years. Close to a million dollars. And no platform.
The session that changed the calculation
The shift happened during a Mindvalley AI Mastery session in late 2025. It was a long, technical, multi-part workshop on n8n automation led by Sara Davison and Tyler Fisk, the co-founders of AI Build Lab.
Emilian was not there for n8n. The thing that stopped him cold was a reference, almost in passing, to Claude Code, which had just become publicly available. “That was my moment,” he says.
What AI Mastery gave him was not just exposure to a tool. It gave him a working understanding of how to build with it:
- How to structure a codebase without a development background,
- How to set up GitHub workflows,
- How to deploy on Railway, and
- How to run QA on software he had written himself.
The reality is, skills don’t come from watching a tutorial. Rather, they come from building something real within a program designed to push you past dabbling.
A massive improvement in terms of patient outcome and financial outcome for ourselves.
— Emilian Popa
Six weeks to build what a million dollars could not
Emilian went home and started. He researched obsessively, spoke to other builders in the community, and taught himself to learn AI through Claude Code professionally. None of it existed in his skill set six months before. Then came Kenya, the elephants, and six weeks of work that delivered what two years and a full development team could not.
Clinician time per patient dropped from seven hours to approximately one. That shift changes the financial model of a functional medicine practice entirely. It changes patient outcomes, too, because a clinician with more time and more energy gives better care.
“A massive improvement in terms of patient outcome and financial outcome for ourselves,” Emilian says.
The platform reads labs and extracts every biomarker. From there, it flags what falls outside functional medicine ranges, cross-references findings against patient history, and generates a draft protocol and health summary. A clinician reviews, adjusts, and approves. All the grunt work is done before they enter the room.
Here is the detail worth sitting with: this is Emilian’s third round of AI Mastery. Each round compounds on the last, and the program keeps pace with where AI actually is, not where it was eighteen months ago.
“The first and second rounds were foundational. The third time was a massive leap,” he says. “Every year is not 40% different. The reality I found this year was actually 90% different.”
What this means for you
Expand Health AI currently has five paying clinics live on the platform, with more in active onboarding across the US and internationally. Emilian built that, without a development team, without outside investment, and without a technical background. He had domain expertise, a clear problem, and a program that showed him how to use AI for business at the level his clinics needed.
And his story isn’t a special case. He’s what happens when someone who deeply understands a problem gets the right tools and the right education at the right moment. That combination is available to you right now.
The gap between what you know and what you can build is smaller than it looks. AI Mastery exists specifically for people at that gap: entrepreneurs, founders, and practitioners who are ready to stop waiting for a developer and start building themselves.
Emilian is already signed up for his fourth round. When the next one opens, the question is whether you will be in it.
The future belongs to the curious. Are you one of them?
