Drafts the move. Waits for your yes.
Sensei proposes a concrete operational move, you approve or refine it, and deterministic code carries it out. The AI is never in the execution path, and every action is logged.
An AI copilot you can actually trust.
Most AI for studios is either a chatbot that only answers, or automation that acts without asking. Sensei is a third shape: it drafts, you approve, code executes, the ledger records. It runs both on a schedule and on demand — on one engine.
Five of your highest-impact moves, already drafted.
Re-engagement, Incident Commander, Renewals, Specials, and Substitute Finder each watch for a specific situation and propose a grounded action — re-engage lapsed members, handle a class incident, catch a failing card, run a timely special, find a sub. You approve, refine, or dismiss.
Ask a question. Get an answer — or a draft.
The Ask Sensei bar runs the same proposer as the agents. Ask how the month went, why a class is soft, who hasn't rebooked — and the answer can turn into a concrete draft you approve. One engine, reactive and proactive.
Trust it earns, not trust you assume.
Every agent starts by asking first. After it has a track record you can see — at least three decisions at 75% or higher approval — you can let it run on its own, and pause it anytime. High-risk moves, like cancelling a class, always ask first, no matter how good the record.
It remembers — and you can make it forget.
Sensei keeps a record of what it has done and what you've set, and you can browse all of it and remove anything. Today that's early — campaign history and your settings — and it's the start of the context that makes Sensei specific to your studio. It's always yours to edit.