OpenAI Developer
Start here if you want to implement and test an OpenAI application.
8 domains About 20 hours suggested
Explore trackKokoAI Academy · Advanced engineering
Three tracks for developers and architects working with ChatGPT, Codex and OpenAI APIs. Study implementation, system design or enterprise architecture.
Explore the tracksIndependent KokoAI Academy training. This is not an OpenAI-issued certification. Public exercises and timed practice provide self-study feedback.
Start here if you want to implement and test an OpenAI application.
8 domains About 20 hours suggested
Explore trackStart here if you want to choose system boundaries and explain design tradeoffs.
5 domains About 13 hours suggested
Explore trackChoose this track if you already design application systems and want enterprise architecture practice.
7 domains About 18 hours suggested
Explore trackEvery domain has source-linked teaching, an inspectable lab, a worked solution, scenario questions with explanations, flashcards and a Koko/Chris discussion transcript. Each track closes with a capstone and public timed practice forms.
Developers should be comfortable reading Python or TypeScript, HTTP and JSON. Architects should bring experience with application design, access controls and operational tradeoffs. Begin with Developer for implementation practice, or Architect Foundations for system boundaries before Professional.
Download the synthetic support-action lab: a starter, reference implementation, fixtures, control tests and an offline evaluation harness. It runs locally with Python and makes no provider calls. Run the downloaded Python file to create a new lab folder with its README, starter and tests. Its README distinguishes tested behavior from production requirements.
Download the lab source (.py)Looking for business workflows? Explore OpenAI for Enterprise Work.