Hamel Husain
Teaches AI evalsPractical writing on how to debug, analyze and measure an AI product, including a guide that answers common questions about evals.
Evals
A set of real inputs with expected outcomes that every prompt or model change runs against. It replaces checking a few examples by eye with a comparison you can repeat.
Building with LLMs Various speakers 1 video
Practical writing on how to debug, analyze and measure an AI product, including a guide that answers common questions about evals.
Detailed, practical posts on building and evaluating AI systems, such as running product evals in three steps.