There will be a chapter on managing the transition. Any specific questions? Generally pilots should be like innovation lab experiments - clear hypothesis, timeline and resourcing to validate or disprove. There will always be learnings in pilots, but then you need to decide whether it's better to extend, re-scope or kill and allocate resources elsewhere. All usual good practices - career development needs to be independent of experiment success or people will only ever join sure bets and high risk, high reward experiments won't get run. Generally these days weekly checkins, small teams, and scoping initial spikes to < 3 weeks (3-6 for bigger projects with intermediate valued deliverables) for most projects. Hardcore rocket science needs longer horizon and more conviction (e.g. building your own foundation model from scratch vs building a harness or a governance layer or context graph).