Learning platform

AI Pipeline Mastery

A course built as software. A Python pipeline generates the course material and executes it as a test, and a Next.js platform runs learners' code either in the browser or on a pool of remote GPUs.

• Built a Python toolchain that generates slide decks, documents and exercises from source. So far it has produced 383 slides and 129 figures rendered from real data.
• Enforced quality by execution. Every reference solution must pass, every starter must fail as expected, and generated files are rebuilt and compared byte for byte before each release.
• Built the learner platform on Next.js 16 with server actions and no internal API layer. The browser never touches the data store, and files are served as signed URLs.
• Designed a GPU job queue for a serverless host. It leases accounts from a pool, tracks jobs through polling and a scheduled sweep, and can run end to end in a local simulation mode.