We embed in one department, map where the hours go, and run a full-day build session where your leaders fix their first problems live.
You leave with a ranked map of where AI pays off for that team, two or three working fixes your team built, and a measured baseline.