Courses

AI coding agents can generate apps from plain English, but they can also make confident mistakes or produce fabricated outputs. In this hands-on workshop, you’ll learn a practical supervise-and-verify workflow to harness AI effectively and reliably. Using an AI copilot, you’ll scope a small project, generate code, and implement checks to ensure correctness and reproducibility, including unit tests, invariants, fixed seeds, provenance checks, and quick sanity plots. You will also practice prompt patterns that reduce brittleness, apply a red-flag checklist for synthetic outputs, and learn best practices for data privacy and API key hygiene. By the end of the workshop, you will be able to decide when an AI agent is appropriate, supervise it end to end, and confirm that results are both accurate and repeatable. Students will leave with a practical checklist, reusable prompt and test templates, a reproducible mini-project, and clear criteria for applying AI agents in product, operations, research, or content workflows.