Not a smarter engine inside a platform — an inversion of the architecture: your profile finally lives on your side, and the market reaches you through your agent.
Built so the market can’t reach you directly — it has to reach your agent first.
Persuasion gives way to verifiable relevance — fit to your state, context, and history, not a bigger ad budget.
Most systems learn from likes, clicks, and purchases. WISP learns from a Decision Session — the moment you’re figuring out what to choose.
Tired, rushed, celebrating, deciding for two — the context you were actually in.
What mattered right now: fast over cheap, quiet over new, safe over exciting.
What you turned down, and why — the signal that leaves no trace inside a platform.
A short check afterward: did it actually fit, or did you wish you’d chosen differently?
We are not building a restaurant app. We’re building the base layer of personal agency — and food proves the loop fastest, because three things line up at once.
Real food decisions happen several times a week — enough repetition to build memory fast.
A mediocre dinner is forgettable. A safe place for an agent to learn and occasionally get it wrong.
Delivery, booking, and pickup already exist. We build only the decision layer.
WISP is pre-MVP on purpose — the first community challenges the product while it’s still shapeable.