To improve clinical success rates with human data

Verge Labs started as Verge Genomics in 2015 with a bet on human data. While most of the field used mice to predict what works in humans, we built a discovery engine that learned directly from multi-modal human data. Over ten years, we’ve built the world’s largest curated patient datasets from CNS tissue and produced an AI-discovered drug that reached the clinic and was tested in ALS patients.

That clinical trial taught us something important. Our first-generation platform could identify new biology and targets, but it missed a critical piece: which patients to drug those targets in. Fortunately, the AI breakthroughs we needed to solve that problem just arrived in the last two years.

In 2026, we became Verge Labs — a frontier AI lab focused on building world models of human disease biology to predict not only which targets to drug, but also which patients to drug them in. We are taking a decade’s worth of proprietary multimodal patient data and hard-won clinical experience to build the next wave of frontier models. The mission is the same: more medicines that work, by reading human biology directly. The execution is different: open the platform to the field rather than develop drugs one at a time.