{"success":true,"filters":{"role":"product","difficulty":null,"search":null},"scenarios":[{"id":"demo-product-prd","title":"PRD Risk Review","role":"Product Management","vertical":"Product","productLine":"AIducation for Product Managers","duration":"9 min","skillAtoms":["Use AI to synthesize research without flattening nuance","Convert ambiguous requests into testable product requirements","Identify adoption, trust, and measurement risks before build"],"toolMissions":["ChatGPT","Claude","Perplexity","Notion AI"],"icon":"🧭","context":"You are a Product Manager reviewing an AI-generated PRD for a feature called \"Smart Reply\" in a B2B customer support product.\n\nThe draft PRD says:\n- \"Agents should get instant AI replies for every ticket\"\n- Success metric: \"increase productivity\"\n- The feature should launch to all customers in 6 weeks\n- It mentions tone customization, but not policy checks, source citations, customer data privacy, or escalation behavior\n- Support leaders have asked for faster first responses, but QA managers are worried about incorrect promises\n\nWrite feedback on the PRD. Your goal is to turn the AI-generated draft into a safer, measurable, buildable product spec.","rubric":{"requirement_quality":{"max":3,"description":"Turns vague requirements into specific, testable product behavior"},"risk_awareness":{"max":3,"description":"Identifies AI product risks such as hallucination, privacy, policy, and escalation failures"},"measurement":{"max":2,"description":"Defines useful success metrics and guardrails"},"prioritization":{"max":2,"description":"Proposes a realistic rollout or scope plan"}},"difficulty":"medium","idealBehaviors":["Replace 'every ticket' with eligibility rules and confidence thresholds","Add policy checks, citations, privacy constraints, and escalation behavior","Define productivity, quality, safety, and adoption metrics","Recommend staged rollout instead of all-customer launch","Include QA manager concerns as launch guardrails","Clarify what users can edit, accept, reject, or report"],"sampleResponses":[{"type":"good","response":"The draft is directionally useful, but it is not ready to build because the core behavior and launch guardrails are too vague.\n\nChanges I would make:\n\n**1. Replace \"every ticket\" with eligibility rules**\nSmart Reply should only generate replies for ticket types with approved policy coverage, sufficient account context, and no open security/privacy flags. Low-confidence replies should route to agent drafting only, not one-click send.\n\n**2. Add missing safety requirements**\nThe PRD needs requirements for source citations, policy checks, sensitive-data handling, and escalation triggers. For example: if the reply mentions refunds, contract terms, account access, or legal commitments, it must cite the approved source or ask the agent to escalate.\n\n**3. Define measurable outcomes**\n\"Increase productivity\" should become:\n- First response time reduced by 20%\n- Agent handle time reduced by 10%\n- QA error rate does not increase\n- Policy-risk flags below agreed threshold\n- Agent acceptance and edit rates tracked by workflow\n\n**4. Roll out in stages**\nStart with two low-risk ticket categories, run an internal beta, then expand by customer segment once QA and support leaders approve the safety metrics.\n\nThis would make the feature buildable while respecting the speed goal and the QA team's concern about incorrect promises."}]}],"counts":{"total":1,"easy":0,"medium":1,"hard":0},"roleCoverage":[{"role":"support","roleLabel":"Support","productLine":"AIducation for Support","firstWedge":true,"demoScenarioId":"demo-angry-customer","scenarioCount":1},{"role":"sales","roleLabel":"Sales","productLine":"AIducation for Sales","firstWedge":false,"demoScenarioId":"demo-sales-outreach","scenarioCount":1},{"role":"marketing","roleLabel":"Marketing","productLine":"AIducation for Marketing","firstWedge":false,"demoScenarioId":"demo-marketing-copy","scenarioCount":1},{"role":"hr","roleLabel":"HR","productLine":"AIducation for HR","firstWedge":false,"demoScenarioId":"demo-hr-policy","scenarioCount":1},{"role":"finance","roleLabel":"Finance","productLine":"AIducation for Finance","firstWedge":false,"demoScenarioId":"demo-finance-analysis","scenarioCount":1},{"role":"engineering","roleLabel":"Engineering","productLine":"AIducation for Engineering","firstWedge":false,"demoScenarioId":"demo-engineering-code-review","scenarioCount":1},{"role":"product","roleLabel":"Product","productLine":"AIducation for Product Managers","firstWedge":false,"demoScenarioId":"demo-product-prd","scenarioCount":1},{"role":"executives","roleLabel":"Executives","productLine":"AIducation for Executives","firstWedge":false,"demoScenarioId":"demo-executive-rollout","scenarioCount":1},{"role":"operations","roleLabel":"Operations","productLine":"AIducation for Operations","firstWedge":false,"demoScenarioId":"demo-operations-sop","scenarioCount":1},{"role":"students","roleLabel":"Students","productLine":"AIducation for Students","firstWedge":false,"demoScenarioId":"demo-student-research","scenarioCount":1},{"role":"teachers","roleLabel":"Teachers","productLine":"AIducation for Teachers","firstWedge":false,"demoScenarioId":"demo-teacher-lesson","scenarioCount":1},{"role":"government","roleLabel":"Government","productLine":"AIducation for Government","firstWedge":false,"demoScenarioId":"demo-government-service","scenarioCount":1},{"role":"healthcare-admin","roleLabel":"Healthcare Admin","productLine":"AIducation for Healthcare Admin","firstWedge":false,"demoScenarioId":"demo-healthcare-admin-message","scenarioCount":1},{"role":"legal","roleLabel":"Legal","productLine":"AIducation for Legal Teams","firstWedge":false,"demoScenarioId":"demo-legal-contract","scenarioCount":1},{"role":"small-business","roleLabel":"Small Business","productLine":"AIducation for Small Businesses","firstWedge":false,"demoScenarioId":"demo-small-business-campaign","scenarioCount":1}]}