How I use AI

You're looking at a site built by a handful of AI agents, on a short leash, working under my direction and review. I'm putting that upfront because it's the clearest evidence I can offer that I do what I coach leaders to do: test AI on my own systems before asking anyone else to trust it.

In coaching, AI touches the parts of the work that happen around a session: prep before we meet, notes after, and materials I build for you between sessions. None of that starts until it's written into our coaching agreement, and nothing that identifies you (your name, your situation) goes into any AI tool before that agreement is signed. Consent comes first, in writing, every time.

Here's what never happens, by design rather than by promise: client data lives in one system built specifically to hold it, separate from everything else I write or work on. It never mixes into my notes on other projects, or into anything I write in public — when something from a session does inform that public writing, it's already been stripped down to a codename and an abstraction, with nothing that traces back to you. That separation is architectural: a property of how the system is built, which holds whether or not anyone remembers to think about it.

If any of that raises a question for you, ask me directly, before you sign anything rather than after. That conversation is one I'd rather have too early than too late.