Outer Biosciences exited stealth at 01:31 UTC on 22 August, four years after being founded in 2022. Its platform keeps surgically discarded human skin alive outside the body for up to 30 days — against an industry norm measured in days — and runs a closed loop in which a model predicts which untested chemicals will act on a specific skin function, the living tissue tests them, and every result, positive or negative, returns to the model.
The company
Four co-founders: Michael Polansky as chief executive, Kyung-Jin Jang as chief scientist, Chris Hinojosa as chief technology officer and Stanley King as chief business officer. Roughly $23m raised from Wing Ventures, Initialized and Hawktail. 19 employees, all but Polansky near Cambridge, Massachusetts. The model runs entirely on-premise rather than in the cloud. About two years went into building the tissue pipeline before the AI loop closed.
What the common framing gets wrong
This will be written up as AI drug discovery. It is not. Polansky says plainly that the company is discovering cosmetic ingredients, which means there is no FDA approval to seek — the route is an industry ingredient name plus OECD safety testing. That is the whole reason the economics work, and it is the most interesting fact about the company. Second, the headline productivity figure — two leads in eighteen months before, roughly one candidate every six weeks after — is a company-reported hit rate, not a success rate. A lead here is an ex-vivo signal in donated skin. None of it is peer reviewed, published or externally replicated. Third, the 30-day claim is softer than the number suggests: Polansky's own words are that day-30 tissue "looks a lot like day-one tissue, but it isn't fully indistinguishable from it."
The scale of the addressable space
There are roughly 120 to 130 FDA-approved over-the-counter skin-drug actives across 13 categories, and about 200 including research-backed cosmetic ingredients. Six leads are currently active, with several dozen further hits logged, and the company judges four of the six likely to reach commercialisation — again, its own assessment.
Where AI-bio money actually lands
The ex-vivo and organ-on-chip field is the practical test of whether better human-tissue models plus AI can displace animal testing. The genuinely scarce asset here is proprietary data that does not exist anywhere to be scraped. But note where a well-funded entrant chose to point it: at the market segment with no regulatory gate. The named competitor, Vivodyne, has raised about $80m.