Back to the lessonPreparing the printout…

Prompt Engineering · Module 1 — AI Chatbots and Task Definition

Period 3 · Define: Problems, Context and Success Criteria

[video] Lecture video — Period 3: Define: Problems, Context and Success Criteria — /api/lesson-media/videos/4d3653ff-a3ca-4f63-bf8e-bb93c553e599-1794840394123-c40c1d4933c1a652.mp4

Flashcards · Period 3

Task brief — A concise agreed description of the problem, audience, evidence and success, in four parts.

Question, context, expected result, success criteria.

Answerable question — A question the available data can actually answer.

Which region has the lowest total net sales?

Metric — A defined quantity used for comparison. "Best" is not a metric until you say which one.

Sum of Net_Sales_Amount per Store_Region.

Constraint — A limit on scope or on what may be inferred.

Do not infer dates, profit or causes.

Output format — The shape of the deliverable.

A ranked four-row table with two observations.

Success criterion — An observable condition that shows whether the result meets the task. It tells you what to fix when it fails.

All four regions appear exactly once.

Evidence and limits — What the data contains and what it lacks. Naming the gaps stops the AI from filling them with guesses.

200 transactions; no dates, costs, stores or currency.

Scope change — A new question the original brief and data cannot support. Change the brief or get more data.

"Which region is most profitable?" when no costs exist.

Prompt Engineering · Period 3 · Define: Problems, Context and Success Criteria · printed on November 21, 2026