
Sign up to save your podcasts
Or


If your VP asked for your AI plan tomorrow, what would you hand them?
In this episode, Erwin walks through a 13-week plan for moving your engineering team from personal AI experiments to a written team policy: four stages, four small artifacts, and one 5-minute action to start today.
Key takeaways
Treat AI adoption like a release, not a hobby. Pick one team, one workflow, and block 3 hours a week for 13 weeks.
Stage 1, Sandbox (weeks 1-3): use AI on real work, try to break it, and keep a failure log.
Stage 2, Staging (weeks 4-6): have AI draft your recurring leadership documents, and review the output with a checklist the way you review code.
Stage 3, Production (weeks 7-10): build a decision ledger, a "where AI fits" grid and one agent runbook card, then pilot it for two weeks.
Stage 4, Operations (weeks 11-13): measure before and after, write a value case with kill criteria, sort uses into risk tiers, and brief your VP in 10 minutes.
Capstone: put it all on one page. That page is version 1 of your team's AI policy.
Your 5-minute action: open a doc called "Team AI runbook card v0" and fill in two lines: what one AI use on your team is for, and what it must never touch.
Read the full issue on tinybigspark.com
From the publisher's feed