A linear, stage-gated pipeline. A compound moves through seven sequential steps — activity, selectivity, ADME, PK, CMC, toxicology, and clinical data — before becoming a drug. Each stage runs on its own, one after another, and has to finish before the next one starts.

Our molecular superintelligence platform integrates advanced AI technologies Enchant, for multimodal endpoint prediction, and NeuralPLexer for protein-ligand structural predictions, with automated, experimental chemistry and biology capabilities as one unified system, guiding design and prioritization of experiments.

The interplay of Enchant, NeuralPLexer and automated laboratory data generation enables us to validate hypotheses and iteratively refine models. This engine for new medicines allows us to make hundreds of novel compounds per week and test them in diverse physicochemical, biological, and metabolic assays.