Today’s outcome

• Plan and execute a scoped QA investigation
• Use AI safely across work surfaces
• Present evidence, uncertainty, and next action
• Identify your apprenticeship track

Capstone cards

• CRM role/permission change
• launch-preview responsive review
• event-processing boundaries and recovery
• small code regression

Use fixtures/capstone-cards.md: public/open-book work, followed by a changed-condition live defense.

The required artifact set

1. Test charter
2. AI task brief
3. Evidence packet(s)
4. Automation proposal or manual rationale
5. Ship/fix/investigate recommendation

Five-minute verdict structure

1. User and business outcome
2. Top risks and approach
3. Evidence and result
4. Limits/unknowns
5. Recommendation, owner, next action

Reviewer questions

• What would disprove your conclusion?
• How did you validate the AI’s contribution?
• Is this a technical symptom, customer impact, or both?
• Why does it block—or not block—a release?

Quality professionals communicate uncertainty

Strong QA does not mean “I found bugs.”

It means “Here is what we know, how we know it, what we do not know, and the safest next step.”

Apprenticeship tracks

• Exploratory/product QA
• Automation and test infrastructure
• Integration/operations QA
• Accessibility and visual QA
• AI evaluation and safety

Your permanent operating method

Outcome → context → constraints → plan → evidence → review → durable playbook

Tools and models will change. This method should not.

Individual reflection

• My strongest QA skill:
• The AI habit I will keep:
• The next skill I will practice:
• The work track I want to explore: