OpenAI Academy Builder Bootcamp: Evals

Evaluations are how teams move from 'it seems to work' to measurable, reliable AI systems, and this OpenAI Academy Builder Bootcamp session focuses entirely on that step. OpenAI deployment engineers Andrew Ginns and Gaurav Kaila cover how to define evaluation criteria, structure datasets, run evals with the OpenAI Evals API, and interpret results to understand system quality.

The session also shows how to combine model-based graders with deterministic checks, spot recurring failure patterns, and compare performance across different configurations. It is aimed at builders who want to replace ad hoc testing with repeatable evaluation workflows. Slides from the bootcamp module are attached to the event page. It is a free livestream; register through OpenAI Academy to attend and access the materials.