Leading AI at Work: Governance, Rollout and Practice / Unit 9 LAI 208 review, 2 to 3 minutes
Capstone Four: Playbook, Rituals, Metrics
The idea
The capstone's fourth part, constructed for this course, writes Priya's domain playbook and sets up the operating model that will keep it alive.
In practice
The utility is not one of the book's six domains, so Priya writes a micro-playbook. Its decision things are the Service Disconnection and the Payment Arrangement, both Gold, because those are what customers dispute and the regulator audits. She traces each of the Disconnection's statuses to a section of the regulator's rules, which is how the medical-exemption status got its clear basis. Her two sentences for AI: it may draft notices, summarize account histories and prioritize call lists; it may never decide whether to disconnect or whether a household is protected. For the operating model, she assigns hats. She is steward for the Disconnection; a writer in the billing team edits the collections Packs; the test lead is matrix custodian; and the architect reviewing vendor tools becomes the AI Safety Lead, with time set aside. The four rituals follow. Object reviews for each Gold thing every six months; one question in the weekly change meeting; the feature gate for every AI proposal; and three lines in the definition of done for collections stories. For metrics, she tracks time to clarity for new officers, rework from misread things, and the override log for the payment-risk model.
Where people slip
The playbook took Priya one afternoon because its moves are the book's moves in every domain. The operating model is what makes it outlast her: if she leaves, the Disconnection still has a steward's hat to hand on, a review in the calendar, and a gate that every AI proposal must pass.
Your place on the trail
Every lesson rests on earlier ones and opens later ones. Tap any to jump.
Builds on
Something wrong in this lesson? Flag it · Sources for Leading AI at Work: Governance, Rollout and Practice