Outer Biosciences is combining living human skin tissue with an AI model in an attempt to speed up the discovery of cosmetic ingredients. The startup, co-founded in 2022 by Michael Polansky, says it can keep donated surgical tissue viable for as long as a month while collecting molecular data from controlled experiments.
The company has operated largely out of public view while building its tissue supply, laboratory system and predictive model. It is now discussing the work as its pipeline expands, although most of the performance figures remain company claims rather than independently reported results.
Living tissue extends the experimental window
Outer Biosciences sources skin that would otherwise be discarded after surgery, mainly plastic surgery, through nonprofit and commercial tissue suppliers. Polansky told TechCrunch that samples are collected under institutional review-board oversight and documented donor consent. They arrive de-identified, with direct personal information removed before the startup receives them.
The company’s proprietary support system supplies nutrients and removes metabolic waste. According to Polansky, this keeps tissue alive for up to 30 days while preserving much of its original structure and molecular activity. Conventional tissue testing often lasts only a few days, which may be enough to examine acute toxicity but not slower processes such as changes in pigmentation, collagen or the skin barrier.
That longer window allows researchers to observe a response over time. Polansky offered UVB exposure as an example: the laboratory can damage a sample and then track inflammatory, stress and recovery signals over subsequent weeks. He emphasized that the team is observing the biology following an injury, not claiming to heal the tissue.
The method is positioned between experiments using simplified cell cultures and testing on people. It still has limitations. Month-old tissue is not completely identical to a fresh sample, and TechCrunch’s report does not cite an independent validation of the platform’s claimed performance.
Experiments feed a closed AI loop
The AI component proposes chemicals that may affect a particular skin function. Researchers then test those candidates on living tissue and return both successful and unsuccessful results to the model. Outer Biosciences describes this as a closed feedback loop that should improve later predictions using data generated by its own experiments.
Before adding AI, the company searched scientific literature and worked with the National Cancer Institute on compounds found in extreme environments. Polansky said that approach produced a couple of leads over roughly 18 months. With the predictive model in place, he said the startup is identifying a new candidate around every six weeks. It currently reports six active leads and several dozen additional hits.
Those figures describe internal progress, not finished consumer products. Four of the six leads appear likely to reach commercialization, according to Polansky, but the compounds must still receive standardized industry names and pass safety testing. Outer Biosciences is discovering cosmetic ingredients rather than drugs, so Polansky said they do not require drug approval from the US Food and Drug Administration.
The company intends to license or sell ingredients to beauty or pharmaceutical businesses instead of launching its own consumer brand. Partners would still need to formulate products, scale manufacturing and manage supply chains. Outer Biosciences also earns revenue from research collaborations, including work on skin rashes caused by some cancer treatments and projects with beauty companies.
Proprietary biology becomes the strategic asset
Outer Biosciences has raised about $23 million and employs 19 people, with all but Polansky based near Cambridge, Massachusetts. Its early backers include Wing Ventures, Initialized and Polansky’s investment firm Hawktail. He did not tell TechCrunch whether the startup is currently raising more capital.
The company faces alternatives including organ-on-a-chip systems, microphysiological platforms and lab-grown human tissue. Polansky argues that its advantage is the integration of real donated tissue, a varied donor pool, controlled conditions and repeated molecular measurements rather than any single technical component.
That combination also creates a proprietary dataset that cannot be assembled by scraping public information. Outer Biosciences runs its AI work on premises, both because its compute requirements are modest and because the company does not want its biological data in the cloud.
The eventual business model is still unsettled. Polansky said the company could remain an ingredient supplier, license discoveries or reorganize around a particularly promising compound. Its immediate ambition is narrower: make the model accurate enough to identify useful directions in skin biology before every possibility has to be tested physically. Whether its experimental loop can consistently turn predictions into safe, marketable ingredients has yet to be demonstrated at commercial scale.
