Delivering & Measuring AI Programs
The program manager's lens on AI: run a rollout across teams and prove it worked — stage gates wired to real signals, a RAID log, and a program scorecard. Runs a real, deterministic scorecard over a three-team rollout, and teaches the discipline that separates a program review from a victory lap: adoption is not success. Read the lagging signals adoption hides.
"Cirrus adopted as well as everyone — and got blocked at GA on override rate and incidents. Adoption isn't ready."
6 Interactive Sessions
Short, interactive sessions — watch it work, steer it, then build it yourself. Go deeper anytime with the full code walkthrough.
- 1
An AI program fails on quality, not the schedule
See that an AI rollout carries risk classes a normal program plan doesn't anticipate — drift, eval decay, over-reliance, incidents — and that quality, not the timeline, is where it fails.
- 2
A gate on adoption alone is theatre
Design stage gates that pair a leading signal (adoption) with lagging quality signals (override rate, incidents), so no team reaches GA on usage alone.
- 3
Name the AI risk, give it an owner
Turn the AI-specific risks into a tracked RAID log — risks, assumptions, issues, dependencies — each with an owner and a mitigation, so nothing important lives only in someone's head.
- 4
Deployed is not adopted (and adopted is not done)
Measure active use rather than seats enabled — and understand that even adoption is only a leading signal, not proof the rollout worked.
- 5
The team that adopts well and isn't okay
Read a program scorecard the way it must be read — leading and lagging signals together — and catch the team whose healthy adoption hides a quality problem.
- 6
Hold the team that adopted well
Make the gate call the scorecard earned — promote the healthy teams, hold the one that adopted well but isn't ready, and report it honestly with a fix, not a date.
Production patterns you'll master
Synthetic data included
- Program plan + stage gates
- RAID log
- Weekly rollout events (3 teams)
- Program scorecard
- Gate verdicts
What you walk away with
Shareable portfolio
A public URL showing your module timeline, patterns mastered, and completion status.
All the code
Download everything as a ZIP — pipelines, guardrails, deployment configs. Yours forever.
Module walkthrough
Each module documented with deliverables and the production pattern you implemented.
Ready to build your delivering & measuring ai programs?
First course free. $20 per course after that.