Wirewalk Instituteof AI
wrk-217 Intermediate Available

AI for PMOs and Portfolio Management

For PMO analysts, portfolio and resource managers, project controls staff and PPM platform owners who must make AI work across a whole portfolio and defend it to a steering committee, an auditor and a regulator.

Level
Intermediate
Study time
16 hours
Lessons
18 in 6 modules
Exam
20 questions

About this course

This course covers the multi-project layer above single-project work (wrk-216). It follows one fictional composite company, Sablewood Systems, a California maker of tactical communications and sensing equipment with 146 active projects. Sablewood has just lost Microsoft Project Online and has had an ERP program report green for seven months before turning red with a $14.2 million overrun. Over six modules you build and test the AI that a PMO actually uses, and you decide where it must not be used. You backtest early-warning rules on held-out history and require an evidence link for every sentence of an AI summary. You use a mixed-integer solver to choose a $150 million portfolio, with the LLM only explaining the result. You design assignment shortlists that meet California's automated decisionmaking technology (ADMT) regulations, including the risk assessment, a human review that is real, the pre-use notice and a disparity check. On a defense program you run DCMA 14-point checks and a Monte Carlo forecast, and you keep variance narratives grounded in the control account manager's (CAM's) notes and CUI inside the enclave. You choose a PPM platform and give its agents least-privilege permissions tested against prompt injection. Finally, you build a capitalization control under ASC 350-40 as amended by ASU 2025-06.

Tools are compared the way PMOs choose them: rules and solvers against LLMs, Copilot against an enterprise LLM API against built-in PPM AI, and Planner, Jira with Rovo, Planview and ServiceNow SPM, each described in the vendor's own published terms, with no vendor performance claims. Exercises use synthetic data and worked outputs, with a read-through path that requires no coding. The 25-initiative selection example supplies its full CSV and solver code; other lesson-scale datasets are not downloadable. The running case is a hybrid of a defense contractor and a commercial product company: module 4 and part of module 3 are defense-specific, and each module 4 lesson says how its method transfers to capital, construction and health-system programs. The capstone is a PMO AI operating plan for the audit committee. It has to answer honestly that PMO AI does not pay for itself in year one or, once internal effort is counted, at steady state, and still win continued funding on the evidence of earlier warning and better decisions. Regulatory and accounting content is educational, not legal or accounting advice. Completion and a passing exam score earn a Wirewalk Institute of AI certificate of completion and assessment. It is not a PMI, vendor or professional certification and carries no PDUs.

What you will be able to do

  • Evaluate a portfolio early-warning system on held-out labeled history (flag precision, recall and lead time with their uncertainty, RAID themes and extracted dependencies) and accept an AI-drafted summary only when every claim has a checked evidence link
  • Analyze AI-assisted demand intake for field accuracy, agreement with the governance panel measured against the panel's own agreement, and length bias, and decide what the panel must still score itself
  • Select a portfolio with a mixed-integer optimizer under budget, skill, mandatory, dependency and exclusion constraints, use an LLM only to explain results and run what-ifs through the solver, and present scenarios with a faithful decision log
  • Evaluate a skill forecast and an assignment-shortlist process for stale data and disparity, and design the risk assessment, real human review, pre-use notice and disparity check California's ADMT regulations require for allocation of work
  • Analyze a defense program's schedule health, Monte Carlo forecast and earned-value position, audit AI-drafted variance narratives for invented causes, and choose where AI may process CUI
  • Evaluate PPM platforms and agents on data fit, license prerequisites, permissions, audit logging, cost and exit, gate agent releases with a prompt-injection test pack, and tier the company's AI uses in an owned inventory
  • Evaluate AI-touched financial numbers: a capitalization control under ASC 350-40 and ASU 2025-06, benefits variance without causal claims, and PMO AI value split into booked, capacity and risk in year-one and steady-state views
  • Create a PMO AI operating plan that sequences use cases by value, data readiness and legal exposure, with evaluation thresholds fixed before testing, controls, owners, monitoring, a tested rollback and an honest value statement

Syllabus

6 modules · 18 lessons · about 16 hours

Module 1: Seeing Trouble Early: Status, RAID and Dependencies Across 146 Projects

Catch green-reported projects the data says are red with backtested rules and evidence-linked summaries, then find systemic risks in RAID logs and dependencies hidden in ticket text.
3 lessons
  1. Catching Watermelon Projects: Rule-Based Cross-Checks and an Evidence-Linked Executive Summary
  2. From 2,310 RAID Entries to Five Systemic Risks: Clustering That Keeps the Owner Link
  3. Dependencies Nobody Linked: Extracting Cross-Program Dependencies from Jira Text into a Graph
  • Decision Lab · 12 minThe Green ReportBuild Sablewood's early-warning portfolio report after Atlas, decide who sees a flag first, and live with your choices when a $1.9 million change order lands on a green project.

Module 2: The FY2027 Portfolio: Intake, Selection Under Constraints and the Investment Committee

Turn 212 funding requests into scored, cited intake records, let a solver pick the portfolio under budget and skill constraints while an LLM explains it, and take a defensible pack to the investment committee.
3 lessons
  1. Turning 212 Requests into Scored Intake Records with Cited Evidence
  2. Letting a Solver Choose the Portfolio and an LLM Explain It
  3. The Investment Committee Pack: Scenarios, a Faithful Decision Log and Records Rules
  • Decision Lab · 13 minThe Cut LineTurn 212 requests into a $150 million FY2027 portfolio, survive the CFO's $135 million fallback and a powerful sponsor, and bring a pack the audit committee can rerun.

Module 3: People, Not Just Capacity: Skill Forecasts and Assignment Shortlists Under California's ADMT Rules

Forecast skill gaps from stale HR data and design assignment shortlists with a real human review, a risk assessment, a pre-use notice and a disparity check under California's ADMT regulations.
3 lessons
  1. Forecasting Skill-Level Demand and Supply When the Skills Data Is Stale
  2. Assignment Shortlists with Reasons and a Human Review That Is Real
  3. ADMT Compliance by January 1, 2027: Risk Assessment, Pre-Use Notice, the Work-Allocation Exception and a Disparity Check
  • Decision Lab · 13 minSixty-Four Open RolesStaff 64 project roles with an assignment-shortlist tool under California's ADMT rules, answer an engineer who asks why she is never picked, and decide what to do about 400 engineers in Ireland.

Module 4: Ridgeline: Schedule Health, Forecasts and Variance Narratives on a Defense Program

Check schedule health, read a Monte Carlo forecast with its calibration, keep AI-drafted variance narratives inside the CAM's notes, and decide where AI may touch CUI on a defense program.
3 lessons
  1. DCMA 14-Point Checks and a Monte Carlo Forecast, with ML as a Labeled Second Opinion
  2. Variance Narratives Grounded Only in CAM Notes, and an EAC That Survives Review
  3. Where AI May Touch CUI: Choosing a Tool Boundary for EVMS Data
  • Decision Lab · 14 minMonth Eighteen on RidgelineDecide where AI may touch CUI on a defense program, report a schedule forecast and an EAC the customer can believe, and answer DCMA's data call about AI in the EVMS process.

Module 5: Running AI Inside the PMO Stack After Project Online

Choose the PPM platform after Project Online, give vendor agents least-privilege permissions gated by a prompt-injection test pack, and find every AI use in the company for a tiered inventory.
3 lessons
  1. Choosing the Post-Project Online Platform: Planner, Jira with Rovo, Planview or ServiceNow SPM
  2. Agent Permissions, Audit Logs and a Prompt-Injection Test Pack
  3. The Company's AI Use-Case Inventory: Discovery, Tiering and Owners
  • Decision Lab · 12 minAgents in the StackChoose Sablewood's platform after Project Online, decide what its agents may read and write, respond to a planted instruction in a comment, and find every AI use in the company. The effort scoreboard adds workstream weeks; it is not elapsed calendar time because configuration and discovery overlap implementation.

Module 6: Money, Benefits and the PMO AI Operating Plan

Build a capitalization control the auditor can test, report benefits without causal overclaiming, and answer honestly whether PMO AI pays for itself, in the operating plan for the audit committee.
3 lessons
  1. Capitalization Classification Under ASC 350-40 and ASU 2025-06 with a Control the Auditor Can Test
  2. Benefits Without Causal Overclaiming, and the Value Case for PMO AI
  3. Capstone: The PMO AI Operating Plan
  • Decision Lab · 12 minThe Q1 CloseAutomate capitalization classification without making the model the control, decide how to adopt ASU 2025-06, report Atlas's benefits, and choose the PMO AI value slide for the audit committee.
  • Capstone simulation · 35 minThe PMO AI Operating PlanRun Sablewood's PMO AI program from the first bets in October to the audit committee in April: sequence six uses, build evidence you own, survive January's pressure and March's scrutiny, and tell the truth about the money.

Exam and certificate

The exam unlocks once you have completed every lesson and played each Decision Lab to an ending. It draws 20 questions from a larger bank, spread across the modules, with the answer options shuffled for each attempt. You have 45 minutes and need 80% to pass. After a failed attempt you can try again after 24 hours, up to 3 attempts in total. Feedback is given per module, never as the answer key.

Passing earns a certificate of completion and assessment from the Wirewalk Institute of AI with a unique ID that anyone can check at institute.wirewalk.com/verify. It is not an accredited qualification or a vendor certification. How certification works.

The Wirewalk Institute of AI is independent. This course teaches publicly documented products and is not affiliated with or endorsed by Microsoft, Atlassian, Planview, ServiceNow, Asana, Smartsheet, Deltek, nPlan, Oracle, or Workday. Product and company names are trademarks of their respective owners.

wrk-110 Basic Waitlist

AI for Finance and Accounting Teams

Practical AI for finance and accounting: reviewing models in Excel with Copilot and Claude, reconciliation and close support, invoice extraction, FP&A commentary, and the SOX and audit-evidence discipline that keeps AI-assisted numbers reliable.

7 h · 15 lessons Join the waitlist
wrk-111 Basic Waitlist

AI for Legal Professionals

Using AI in legal work without ethical or evidentiary missteps: ABA Formal Opinion 512 and the Model Rules, grounded research in Westlaw and Lexis, contract review, eDiscovery, court AI-disclosure orders and building a practice-group AI policy.

7 h · 15 lessons Join the waitlist
wrk-112 Basic Waitlist

AI for HR and Talent Teams

AI across the employee lifecycle with the legal guardrails built in: drafting and screening in recruiting, NYC Local Law 144, Illinois HB 3773 and the EU AI Act's employment rules, handbook assistants, performance writing, and employee privacy.

7 h · 15 lessons Join the waitlist