There are now hundreds of startups applying machine learning to healthcare. Drug discovery. Clinical workflows. Diagnostic imaging. Patient engagement. From a distance, they all sound similar. Up close, the difference between the winners and the noise comes down to one thing: who owns the data.
The companies that build lasting value in AI healthcare are the ones sitting on proprietary datasets — unique, defensible collections of clinical, genomic, imaging, or operational data that make their models smarter over time and impossible to replicate. That's the flywheel. That's the moat.
This isn't a thesis we read in a report. It's one we built by reviewing the earliest-stage AI healthcare pipeline in North America — company by company, deck by deck, year over year. We know which sub-niches are overcrowded, where founders keep solving the same problem with the same approach, and where massive markets are still dramatically underserved. That's where we invest.
Proprietary data is the only durable moat in AI healthcare. Full stop. This isn't a hypothesis — it's the filter we run every company through.
We don't just write checks. We study the landscape obsessively and publish what we find. The fund is a byproduct of the knowledge.
We see companies larger funds structurally miss. Early. Often. Before the market catches up. Our pipeline grows by 20–30 new companies every week.
Four verticals. US and Canada only. AI must be core. We go deep where others go broad — which means we understand the regulatory, clinical, and commercial dynamics well enough to be a useful partner.
Most emerging managers ask you to take their thesis on faith. We published ours. The State of AI Healthcare is a 60+ page annual report built from our proprietary dataset — the same intelligence that drives every investment decision. We'd rather you read our research than listen to a pitch.
Get the Report →We don't just label sectors — we have a point of view on each one. Here's how we think about the space, and what we're looking for.
The deepest moats, the longest timelines, and the highest ceiling. PhD-led teams with proprietary datasets thrive here. Everyone else struggles.
Deep Moats, High CeilingThe most underbuilt vertical in AI healthcare relative to market size. If you're building here, we want to hear from you.
Most Underbuilt VerticalHardware-AI integration is the next frontier. Regulatory complexity filters out the casual entrants and rewards the ones who persist.
Next FrontierThe largest vertical in our pipeline — and the most crowded. Winning here requires specialty-specific workflows or a novel go-to-market, not another general-purpose platform.
Crowded — But Room for BreakoutsWe do not invest in companies where AI is a wrapper, a workflow enhancement, or a chatbot UI. We look for products where removing the AI would break the product entirely. Machine learning, computer vision, foundation models — doing the fundamental work the product promises.
We define this precisely: an operational corporation already established, with team members physically based in either country. We do not fund companies "planning" to establish in the US or Canada. You must already be operational there.
Biotech/pharma tools (no therapeutics), diagnostics, medical devices, or healthcare SaaS. We go deep, not wide — which means we understand the regulatory, clinical, and commercial dynamics of each vertical well enough to be a useful partner, not just a check writer.
We back founders with fast $50K checks at the earliest stages, when the product is still being shaped and the team is small. We don't string founders along. We move fast, give clear answers, and get to yes or no quickly.