Our Story
Brodsky began as a side project from a frustrated neuroscientist, Ben Buller. Tired of slow, expensive discovery cycles, Ben started hacking nights and weekends—mixing wide-space generative chemistry with rigorous scoring—until he stumbled into a faster, saner way to find drug ideas.
What emerged wasn't another black box. It's a diversity-first engine that proposes synthesizable molecules, ranks them with Boltz-2, and (when useful) adds pose-aware SAR for clarity. It works across prokaryotic and eukaryotic targets—cofactors, transmembranes, the messy real world—without forcing dogma like hard Lipinski gates.
What we believe
- Explore widely, then prioritize. Optimization comes after coverage.
- Explain the "why." Scores, poses, and literature context—no mystique.
- Keep IP yours. Run locally or in a locked-down cloud.
- Pragmatism over purity. Heuristics guide; data decides.
Brodsky exists to shorten the distance from target to testable—so you spend less on guesswork and more on what works.
