# AI PM Skills and Portfolio Scorecard

Use evidence levels rather than self-awarded expertise labels:

0 = no artifact; 1 = practice with fictional material; 2 = reviewed work with real approved context; 3 = applied work with observed outcomes and clear personal contribution.

| Capability | Evidence to show | Your level and link |
|---|---|---|
| Problem framing | User task, evidence, alternatives, and unknowns | [ ] |
| Research judgment | Source-linked synthesis and contradictions | [ ] |
| Prioritization | Tradeoff, uncertainty, and review trigger | [ ] |
| Prototyping | One testable journey and honest simulation boundaries | [ ] |
| Technical collaboration | Feasibility discussion, permission model, reviewed handoff | [ ] |
| AI evaluation | Cases, rubric, failures, and release reasoning | [ ] |
| Measurement | Defined denominator, baseline, and limitations | [ ] |
| Communication | A decision update a stakeholder can understand | [ ] |
| Creator capability | Audience access, validation, distribution, operating model | [ ] |

Do not sum these into a universal employability score. Select the capabilities relevant to your intended work and choose the next missing artifact.

## Case study outline

1. Context and your role.
2. User problem and evidence.
3. Alternatives and the consequential decision.
4. Prototype, workflow, or implementation.
5. What AI did; what you decided, checked, and corrected.
6. Evaluation or research method and limitations.
7. Observed result, with source and time window if available.
8. What you changed and what you would investigate next.

## Integrity check

- [ ] Practice work is labeled as practice.
- [ ] Simulated numbers and fictional users are labeled.
- [ ] Employer/client material has appropriate permission and redaction.
- [ ] Team outcomes are distinguished from personal contribution.
- [ ] No invented metrics, credentials, research, or tool experience.
- [ ] Production readiness is not implied by a prototype.

## Two-minute explanation

The user needed [ ]. I learned [ ] from [ ]. We chose [ ] instead of [ ] because [ ]. My contribution was [ ]. AI helped with [ ], and I corrected [ ]. The result was [observed outcome or learning]. Next I would [ ].