Thesis

The category

High-performing adults spend on self-development in three disconnected buckets: content they mostly don't finish, one-to-one coaching they can rarely afford weekly, and rooms — masterminds, retreats, dinners — they join for the other people in them. Nothing connects the three. The content has no memory of you, the coach has no visibility between sessions, and the room disperses on Sunday night.

The wedge

We start with the moment of avoidance, not the calendar. The wedge is an always-available coach that holds one continuous thread about a single person's stuck point, plus a small physical room of people who know that thread. The dinners come first because they are cheap to run, impossible to fake, and they generate the transcripts that make the coach worth having.

Why now

Two shifts. Models are finally good enough to hold context across months and push back without sounding like a customer-service script, which is what separates a coach from a chatbot. And this cohort has priced traditional weekly coaching out of reach at exactly the point where they have the most to gain from it.

What we're testing right now

Pilot 001 is 25 sessions run by hand, free, by the founder. We are testing four things: whether people can name a crux in 25 minutes; whether a specific if-then commitment survives seven days; whether the challenger or supporter voice produces more follow-through; and whether dinner attendees convert into people who want the coach between dinners.

What would falsify it

If sessions produce a named crux but no behaviour change at the seven-day check-in, the diagnostic is entertainment and the thesis is wrong. If dinner attendees enjoy the room but never ask for anything between dinners, this is an events business, not a product. If people will only engage when a human is on the other end, the economics don't hold and we should say so rather than dress an agency up as software.

What we won't do

No clinical claims, no outcome claims we haven't measured, no engagement mechanics designed to make the product feel necessary. If the honest version doesn't work, we'd rather find that out early.

Reach us

lilykang27@gmail.com. Disagreement is more useful than interest — the changelog shows what people have already changed.