The bedrock
Version · 24 June 2026
This thesis evolves as I get real industry exposure. Earlier dated versions live in the Thesis menu above.
Why I'm doing this, and what the durable edge actually is.
The durable edge is clinical and scientific depth — then compounding it with the right business models, relationships, insider access, a domain regulated enough that depth is scarce, and a technical co-founder who can build.
Clinical and scientific knowledge is the foundation — and it keeps deepening through immersion: reading, and talking to the people I know from med school who are still inside it. But on its own it's necessary, not sufficient. The edge is that foundation plus four things: fluency in business models — what's investable, what's been tried, where the money moves; relationships with the people who buy, deploy and decide; insider access to the institutions; and depth in a domain regulated enough that the depth is rare. The knowledge is the floor; the terrain is the edge.
When the marginal cost of building software falls toward zero, "we built it" stops being a moat — anything a small team ships in a weekend, another rebuilds the next. Value migrates to whatever doesn't get cheaper. In consumer and non-regulated SaaS what's left is thin: distribution, brand, switching costs. In healthcare it's a thick stack cheap code can't touch.
This is the shift YC named in "The Age Of The 40-Year-Old Solo Founder Is Here": when code is nearly free, the scarce thing is knowing what to build and having the standing to build it where it's hard — even if my read is that the domain half still pairs best with a technical builder, not a true solo run.
As building gets free, moats are what's left.
| Regulatory clearance | MHRA SaMD classification, DCB0129/0160 clinical safety, DTAC — earned, not coded. |
| Clinical evidence | An outcomes base a rival needs years and real deployments to replicate. |
| Data governance & access | IG, Caldicott, DPIAs, data-sharing agreements — plus proprietary data with a compounding loop. |
| Workflow entrenchment | Integration into a clinical workflow that's painful and risky to rip out. |
| Liability & trust | Risk-averse institutional buyers who move on credibility, not features. |
As the build moat goes to zero, every row's relative value rises. The most regulated, relationship-gated, evidence-heavy industry there is, entered as a credentialed insider, is a good place to stand.
Domain expertise is not the moat. It is what lets me see and build the moat.
Plenty of people know healthcare deeply; knowledge in my head is table stakes, not defensibility. It earns its keep by letting me spot the defensible wedge outsiders can't see, then build the clearance, the data loop, the sticky integration, the evidence. My depth lets me find and build moated things — it is not itself the moat.
And it's being refreshed at the right moment. Healthcare has changed enormously since I left med school; re-entering in an AI-native era both deepens what I know and lets me re-read the field through a commercial lens I didn't have before — built through Studdle, Decrey and EF. On EF selection day the signal was consistent: people came to me for the health background, how much I knew clinically and scientifically. That's the raw material; the commercial layer turns it into an edge.
The access isn't the kind consulting rents you for a project: knowing people in US payer strategy, directors of operations inside the NHS. Health tech demands dual vision — you underwrite health ROI and commercial ROI at once. Health ROI I read natively, because I was trained to. Commercial ROI needs a different lens and different connections, which immersion builds. The thesis is the marriage of the two.
Dual vision: native health ROI × acquired commercial ROI.
Picking the wedge rewards breadth and evaluative sharpness — what consulting and a wide commercial map give you. Building the moat is institutional-deployment craft: a safety case through redraft, NHS procurement, integrating against a real trust's EHR, standing up the evidence. That craft only forms against a real institutional counterparty — founding, operating, or hired onto the team. It can't be a side project, because a side project has no institution on the other side. Both muscles matter; they're built in different places.
Different muscles, built in different places.
Regulation raises the floor and slows everyone equally — it doesn't hand me an edge. Babylon had maximal insider access in the most regulated industry there is and still collapsed: access won contracts ahead of a validated product and sustainable economics. "Hard to enter" is not "moat." The moat is the specific thing I build.
It comes with a tax I'm choosing knowingly: longer cycles, heavier capital, a heavier evidence burden, slower iteration. Immersion lets me watch others pay it — how they hold cash across an eighteen-month cycle and pitch that timeline; how they actually assemble an evidence base. I can research the mechanics online. What research can't give me is the relationships, and seeing it happen rather than reading about it. Same learnings, different weight.
The winning pattern isn't the lone clinician — it's domain credibility plus a technical/operator complement. Across 300+ health-tech unicorn founders, a slim majority were insiders, but the through-line was teams (80%+) pairing credibility with build capability, and seniority (70% had ten-plus years). My job is to be the credible domain half and pair deliberately. That search runs constantly.
I commit to health as the 10–15 year domain on defensibility grounds, not preference. I use whichever vehicle I land — consulting or commercial — to sharpen the pick while I find the technical co-founder and the validated wedge that let me build. The institutional craft I lack isn't a blocker: it's learnable inside my own company with the right co-founder and a real design partner, or hired. Founding stays the destination; the two paths are just the ways in, and both keep it live.