Heuresia

Heuresia is an early-stage company applying machine learning to drug discovery - starting with the question that kills most drug candidates before anything else does: is it safe? Most compounds fail not because they don't work, but because they turn out to be toxic, often after years and millions of dollars invested. We're building a system that predicts that risk computationally, before a molecule is ever synthesized. We're pre-seed and pre-funding. Right now, that means a small founding group building toward one clear proof point: a toxicity-prediction model that outperforms the best publicly available tools on real benchmarks. Everything else from molecule generation to full simulation of drug-protein interactions is the direction we're building toward once that foundation holds. We're not a large company yet. We're a handful of people: ML engineers, biochem and pharma-adjacent researchers, and backend engineers - building this from the ground up. If you want to know more about where this is headed, reach out, we're happy to share the roadmap.