I work as the Head of AI Evals at Every, helping the team build benchmarks to professionally test new models from all the major labs, as well as designing the system that optimizes skills for our self-improving AI agent product. I have been an AI engineer since the GPT-3 beta in 2020, and have written two books for O'Reilly on Context Engineering with DSPy and Prompt Engineering for Generative AI. Over 400,000 people have taken my AI courses on Udemy. Previously I co-founded and built a 50 person growth agency, responsible for operations, data science, and technology teams.