Summary
Jason Madeano is a machine learning engineer and founder with eight years of experience building production recommendation systems, large language models, and reinforcement learning solutions across industry and MIT research. He helped scale ML features at Pinterest and previously advanced Theory-Based Reinforcement Learning at MIT, earning a Best Paper award at the NeurIPS Cooperative AI Workshop for work that combines probabilistic program induction with explainability. Jason’s research background includes designing large-scale human experiments and symbolic message-passing methods to improve knowledge transfer between agents and people, and he has applied state-of-the-art NLP techniques like GPT fine-tuning to bridge human and model behavior. Now based in San Francisco, he’s co-founding NextByte to help people take control of their web, bringing a rare mix of rigorous academic modeling and hands-on productization experience. Notably, he holds both a Bachelor's and an MEng from MIT with near-perfect marks, reflecting deep technical training paired with entrepreneurial drive.
8 years of coding experience
6 years of employment as a software developer
Master of Engineering - MEng, Computer Science, 5.0/5.0, Master of Engineering - MEng, Computer Science, 5.0/5.0 at Massachusetts Institute of Technology