Michael Polansky Is Training AI on Living Human Skin

The company sources surgical discards through vetted biobanks, ensuring samples arrive de-identified and viable. By sustaining this tissue well beyond the industry-standard window of a few days, Polansky’s team can observe slow-motion biological processes—such as collagen remodeling and inflammatory responses—that were previously invisible to rapid testing. This pipeline feeds a proprietary AI system that identifies promising cosmetic ingredients, currently generating new candidates every six weeks.

While competitors like Vivodyne are also leveraging organ-on-a-chip technology, Outer Biosciences maintains a distinct approach by focusing on real human tissue rather than synthetic models. The startup has raised roughly $23 million to date and employs 19 people, mostly based near Cambridge, Massachusetts. Rather than launching a consumer brand, Polansky intends to license these data-backed, safety-tested ingredients to established beauty and pharmaceutical companies. For Polansky, the goal is to shift dermatology away from trial-and-error discovery toward a predictive model that could eventually map skin biology with unprecedented precision.