We just hit a milestone that made us stop and reflect. Three years ago, we made an audacious bet: that you can treat automated chemistry as a learning problem — and solve it by scaling chemical data generation.
