AI Product Management
Spec the probabilistic, price the variable, ship the uncertain
Discovery at the capability frontier, evals as specs, trust-calibrated UX, unit economics, and the metrics that make AI products work.
Module
Discovery in the AI Era
Feasibility is the unknown variable — start from what models can actually do
Module
Spec'ing the Probabilistic
Evals as requirements, confidence as a product surface, and shipping behind gates
Module
The Business of AI Features
Metrics that survive Goodhart, model decisions, pricing, and the PM's operating loop
Capstone project
Spec an AI feature
Write the product spec for an AI feature: user need, model requirements, failure handling, and the metric that defines success. Assume the model will sometimes be wrong.
Deliverable: A PRD-style spec with an error-handling section and measurable success criteria.
Milestones
Finish every lesson to earn the AI Product Management Certificate.