10x Science × YC

10x Science is building AI-native infrastructure for protein characterization, the molecular-level analysis that pharma teams run on every antibody, cell therapy, and engineered protein before it can move toward the clinic. Today that work falls to PhD-level scientists who sift through tens to hundreds of thousands of files per sample and fill in results through manual curation, a process that can take weeks, months, or even a year per candidate. The company's platform ingests gigabytes of raw instrument data, coordinates AI agents to identify what proteins are in a mixture and how they are structured, and returns reports in minutes with explainable traces that scientists can steer and audit. The launch matters because AI-powered drug discovery is generating more candidates than development pipelines can evaluate , and characterization has become the chokepoint between a promising hit and a viable clinical program. The target users are biopharma R&D teams working on biologics, antibodies, and cell therapies who need reproducible, enterprise-ready output without growing headcount linearly with their candidate volume. The founders are David Roberts and Andrew Reiter, both biochemists, and Vishnu Tejus, a repeat founder with a computer science and AI background , a team that came together out of Stanford. They are announcing a $4.8M seed led by Initialized Capital alongside the launch, and are hiring as they expand the platform.