AI for Project Managers
Every case follows the same loop: a worked example, an assignment you complete on realistic sample data, and AI review against the rubric below. Each case takes 30–45 minutes. This is the working syllabus for the founding cohort (early 2027); founding members can suggest changes.
Module 1. Project Kickoff: Charters, Scope & Plans in Hours, Not Days
Draft a project charter from a messy brief
A VP sends a 2-paragraph email and a call transcript; you need a charter (goals, scope, out-of-scope, success metrics, stakeholders) for Friday's kickoff.
Your assignment
A one-to-two-page charter plus a list of 5 clarifying questions for the sponsor.
What the AI reviewer checks
- Objectives are SMART and tied to business outcome
- Scope and out-of-scope explicit
- Success metrics measurable
- Assumptions and constraints listed
- Clarifying questions target real ambiguities (AI gaps flagged, not filled with invented facts)
Tools
ChatGPT/Claude, Copilot in Word, meeting transcripts
Build a WBS and first-cut schedule
Launch of a new customer portal: you need a work breakdown structure, dependencies and a rough timeline.
Your assignment
WBS (3 levels), dependency list and a Gantt/timeline (Excel, MS Project, Smartsheet, Asana or Jira import CSV).
What the AI reviewer checks
- 100% rule: WBS covers full scope, no overlaps
- Dependencies logical, critical path identified
- Estimates include ranges or buffers with rationale
- Deliverables, not activities, at level 2
- Import-ready format for the team's tool
Tools
ChatGPT/Claude, Copilot, Jira/Asana/Smartsheet AI
Module 2. Meetings & Communication That Run Themselves
Meeting notes to decisions and action items
You have a 60-minute steering committee transcript; stakeholders want decisions, owners and dates within the hour.
Your assignment
Structured minutes: decisions, action items (owner/date), open questions, risks; plus the follow-up email.
What the AI reviewer checks
- Every action has owner and due date
- Decisions separated from discussion
- Nothing invented that isn't in the transcript
- Follow-up email is under 200 words and skimmable
- Sensitive remarks handled appropriately
Tools
Copilot in Teams, Zoom AI Companion, ChatGPT/Claude
Weekly status report for executives
Every Friday you compile updates from 6 workstreams into an exec status report with RAG status.
Your assignment
An exec status report (1 page) generated from raw updates, with a reusable prompt/template.
What the AI reviewer checks
- RAG status justified with evidence
- Leads with what execs need to decide or know
- Risks/issues have mitigation and owner
- Consistent format week to week
- Under one page
Tools
ChatGPT/Claude, Copilot, Jira/Confluence AI
Handle a difficult stakeholder message
A sponsor emails angrily that the timeline slipped; you need a calm, factual reply that resets expectations.
Your assignment
Your reply draft (with AI), the prompt used, and a note on what you changed from the AI version and why.
What the AI reviewer checks
- Acknowledges concern without over-apologizing
- Facts: cause, impact, recovery plan
- Clear options/next step and decision needed
- Tone appropriate for seniority
- Learner edited AI draft for authenticity
Tools
ChatGPT/Claude
Module 3. Risk, Change & Decision Support
Build a risk register and pre-mortem
Before a CRM migration go-live, leadership asks: what could go wrong?
Your assignment
Risk register (risk, probability, impact, owner, mitigation, trigger) with at least 12 risks from an AI-facilitated pre-mortem.
What the AI reviewer checks
- Risks are specific to this project (not generic)
- Probability/impact scored consistently
- Mitigations actionable with owner
- Triggers/early warning indicators defined
- Learner added risks AI missed
Tools
ChatGPT/Claude
Change request impact analysis
The client wants to add SSO integration mid-project; you must assess impact on scope, schedule, budget.
Your assignment
Change request assessment with options (accept, defer, trade-off) and recommendation.
What the AI reviewer checks
- Impact quantified for scope, time, cost, risk
- At least 2 options compared
- Recommendation justified
- Stakeholder approvals path stated
- Assumptions explicit
Tools
ChatGPT/Claude, Copilot in Excel
Module 4. Data-Driven Project Control
Analyze a Jira/Asana export for delivery risks
You export 400 tickets; you need to find bottlenecks, aging items and velocity trends before the QBR.
Your assignment
Analysis with 3 charts (cycle time, aging WIP, throughput) and 5 insights with actions.
What the AI reviewer checks
- Metrics defined correctly
- Charts readable and labeled
- Insights tied to data, not generic agile advice
- Actions specific (who/what/when)
- Data limitations noted
Tools
ChatGPT/Claude (data analysis), Copilot in Excel
Budget tracking and earned value snapshot
Finance asks whether the project will finish on budget; you have planned vs actual costs and % complete.
Your assignment
EVM snapshot (PV, EV, AC, CPI, SPI, EAC) with a short interpretation.
What the AI reviewer checks
- Formulas correct
- Interpretation correct (CPI<1 means over budget, etc.)
- EAC method stated
- Recommendation for corrective action
- Readable for non-PM audience
Tools
ChatGPT/Claude, Copilot in Excel
Module 5. Lessons Learned & Your AI PM Playbook
Retrospective synthesis and your AI impact brief
Project closed; you have survey comments, retro notes and metrics, and want to show leadership what you delivered and how AI helped.
Your assignment
Lessons-learned report (themes, root causes, recommendations) + a 1-page personal AI impact brief with before/after time log.
What the AI reviewer checks
- Themes grounded in actual comments
- Recommendations actionable for next project
- Time savings measured
- Follows company AI policy
- Brief makes a concrete career ask
Tools
ChatGPT/Claude/Copilot