Insitro

Insitro combines large-scale biological data generation with machine learning to accelerate drug discovery, founded by Daphne Koller with the thesis that ML can transform pharmaceutical R&D when trained on purpose-built datasets rather than historical clinical data. The company generates its own training data through automated cell biology experiments, creating a tight feedback loop between wet-lab measurements and computational predictions. Engineering challenges span high-throughput image analysis, causal inference on biological datasets, and building ML infrastructure that handles the unique characteristics of experimental biology data including batch effects and high dimensionality. Their hiring patterns emphasize the convergence of ML engineering and biological sciences, with roles that require both production software skills and understanding of experimental design.