Polyphron, a two-year-old New York City biotechnology company, said Tuesday it has closed a $20 million seed round led by Quiet Capital to build what it calls a verification layer for artificial intelligence in drug development.
The financing, which includes backing from Gradient Ventures, Haystack and Compound, reflects growing investor interest in companies attempting to ground AI-driven drug discovery in actual human biology. Earlier backers include Humba Ventures, Boost and the founders of Chai Discovery and Loyal.
Polyphron trains machine learning models on induced pluripotent stem cell-derived human tissues, testing hypotheses that computational biology systems generate. The company's premise: AI alone won't deliver breakthrough therapies without a scalable way to validate predictions against living cells.
"The missing piece is a scalable environment where models can test predictions against living human biology," CEO Matthew Osman said in a statement, though he might have understated the challenge. Manufacturing human tissue at industrial scale remains an unsolved problem across the biotech sector.
The seven-person team operates from Alexandria LaunchLabs in upper Manhattan. Osman previously co-founded Legit, a machine learning platform for intellectual property disputes in life sciences, and was involved in another venture called Treat AI. Fabio Boniolo, the chief scientific officer, comes from stints at Dana-Farber Cancer Institute, the Broad Institute and Harvard Medical School.
In early September, the company added Vinh Q. Tran as chief AI scientist. Tran spent time at Google DeepMind working on post-training methods and self-improvement systems for Gemini. George Pilitsis, now chief operating officer, previously founded and scaled the Datapoints product line at Ginkgo Bioworks.
Gradient Ventures wrote in a blog post that the capital will fund expansion of Polyphron's manufacturing facilities and refinement of what it calls "tissue world models." The company currently produces cardiac tissue across multiple donor lines and plans to introduce liver tissue models for drug metabolism and toxicity testing.
Its core technology, dubbed the "Tissue Foundry," automates micro-tissue production from stem cells by mimicking developmental biology. A companion AI layer simulates how tissues respond to genetic, chemical and environmental perturbations, training on longitudinal data from donors with varied backgrounds.
Polyphron has partnered with Cellino to construct what both companies describe as an end-to-end personalized tissue pipeline. Under that arrangement, Cellino provides patient-derived iPSCs while Polyphron handles tissue engineering. Published results reportedly demonstrated reproducible cortical tissue architecture across donor lines without extensive re-optimization, though independent validation of such claims typically takes years.
The startup enters a field populated by organ-on-chip developers including Emulate Bio, Novoheart, CN Bio and Hesperos, all attempting to build in vitro human tissue systems for preclinical testing. Emulate recently submitted qualification materials for its Human Liver-Chip to the FDA's ISTAND program, marking incremental regulatory progress for the broader category.
More than 100,000 people remain on U.S. organ transplant waiting lists, a persistent gap despite advances in tissue engineering. Michael Bloch, a partner at Quiet Capital, framed the investment thesis plainly: "If you can make tissue a manufacturable, testable unit at scale, you change the economics of drug development."
Whether Polyphron can deliver on that vision remains an open question. The company emerged from stealth recently with coverage in Core Memory and an appearance on the Free Radicals podcast, where Osman outlined ambitions to "produce human tissue from iPSCs at scale and at incredibly high margin." He positioned the technology as both a validation substrate for AI-generated drug candidates and a potential source of immune-compatible grafts for therapeutic use.
The company has advertised positions for scientists with expertise in iPSC work and tissue engineering at its Long Island City operations, signaling plans to expand headcount as manufacturing ramps. For now, Polyphron is betting that the pharmaceutical industry's appetite for better preclinical models will justify the investment required to industrialize living tissue production.
