Summary
Ari Morcos is a founder and machine learning researcher who now leads DatologyAI, building tools to help teams train higher-quality AI models by improving data. He brings 11 years of deep research experience from Meta AI Research and DeepMind, where his work uncovered mechanistic insights into neural network computation and the role of data in shaping useful representations. Ari’s research earned Outstanding Paper awards at NeurIPS and ICLR, and his background in experimental neuroscience (Ph.D. from Harvard) informs a rare cross-disciplinary approach to model interpretability and behavior. He has translated neuroscience-inspired methods into practical ML diagnostics and data-centric techniques that reveal why some solutions generalize better than others. Based in Redwood City, he combines academic rigor with startup execution, focusing on actionable insights that move ML systems from theory into production.
11 years of coding experience