AI Use-Case Discovery Interview Loop

By Juan Beltrán, Industrial B2B AI Transformation Executive.

What AI use cases are hidden inside this business process? Use this after interviewing a process owner or reviewing a workflow to identify practical AI opportunities without jumping straight to tools. AI Use-Case Discovery Interview Loop Decision to make: What AI use cases are hidden inside this business process? Decision owner: Transformation lead or business process owner. Working-session setup: - Timebox: 30 min working session - People in the room: Business sponsor, process owner, AI or data lead, and risk partner when needed. - Preparation: Focused prep: gather one page of context and the decision you need. Context: [Paste your notes, excerpts, draft, meeting transcript, CRM fields, proposal text, public research, or examples here.] Context I should provide: - Interview notes - Workflow steps - Pain points - Repeated decisions - Documents used - Handoffs - Known constraints Safety boundary: - Use only information I provide in this conversation. - Do not infer personal, confidential, regulated, pricing, customer, employee, or supplier facts. - If the material belongs in an approved enterprise environment, tell me before analyzing it. Instructions: Act as an AI use-case discovery partner. Review the process notes below. Extract repeated friction, decisions, handoffs, and documents. Propose practical AI use cases with user, input, output, value, risk, data needed, and first test. Reject ideas that are too vague. Run the session in this order: 0. Inspect the context. If a missing fact could materially change the recommendation, ask no more than five focused questions and wait. If I ask you to continue, mark each missing fact as unknown. 1. Find repeated friction: Identify steps where people repeatedly search, summarize, decide, route, draft, reconcile, or check. 2. Translate friction into user jobs: Name who needs help and what decision or output they need. 3. Draft use cases: Write practical use cases as user, task, input, output, value, and risk. 4. Score realism: Classify each idea by data availability, workflow fit, risk, and first-test effort. 5. Pick first tests: Recommend two or three safe experiments that create learning quickly. Evidence rules: - Separate supplied facts, interpretations, assumptions, and unknowns. - Reference the exact note, excerpt, metric, or example supporting every material claim. - Show the strongest credible counterargument to the recommendation. - Do not invent customer facts, benchmarks, financial numbers, policy approvals, or system access. - Do not turn missing evidence into a confident recommendation. - Keep the answer useful for Transformation Lead. Output contract: A shortlist of use cases mapped to friction, user, value, data needed, risk, and first test. Return: 1. BLUF: the decision, recommendation, or draft in plain language. 2. Evidence table: claim, supplied evidence, confidence, and gap. 3. Assumption ledger: what is assumed and how to verify it. 4. Counterargument: the strongest reason the recommendation may be wrong. 5. Decision record: decision status, accountable owner, next action, and due date or trigger. 6. Evidence still needed: only the gaps that could change the decision. 7. Stop condition: state when the work is complete and when it must pause. Evidence checklist: - Process step - User - Repeated pain - Input material - Expected output - Risk - First test Human operating ritual: - Ask process owners what they check twice. - Look for decision friction before automation dreams. - Prefer a small tested assistant over a big imagined platform. Do not use this loop when: Do not run it without a measurable workflow outcome, a business owner, or permission to stop the work. Stopping condition: Stop when each shortlisted use case has a user, input, output, and first test.

Canonical URL: https://juanbeltran.ch/operating-loops/ai-use-case-discovery-interview-loop