Florentin Purcea

Built, awaiting client goAgency product, piloted on a dental clinic

An AI-search visibility operating system

Poor AI visibilityBad tracking

  1. Issue detected

    Nobody could say, with numbers, how AI assistants describe the brand — or whether a content change moved anything.

  2. Verification

    AI engines sampled through official APIs; every rate reported with its sample size and a Wilson interval; brand-truth facts checked against what the AI actually claims.

  3. Root cause

    AI visibility was judged by anecdote — one prompt, one day — with no sample, no baseline and no attribution back to traffic.

  4. Fix

    A closed loop — discover, measure, diagnose, prioritise, fix, validate, learn, attribute — with opportunities turned into briefs and experiments, GA4 AI referrals attributed, weekly runs on a durable scheduler and the client report archived.

  5. Measured Delivered

    13 MCP tools · 2 approved opportunities executed on the pilot

    Window: staging, piloted on a dental clinicCaveat: Nothing is shown under n < 5 — by rule. No visibility uplift is claimed yet.

What was built

The system, in buyer words.

A closed loop — discover, measure, diagnose, prioritize, fix, validate, learn, attribute — sampling AI engines through official APIs, every rate with a sample size and a Wilson interval, brand-truth facts checked against AI claims, opportunities turned into briefs and experiments, GA4 AI referrals attributed, weekly runs on a durable scheduler, client report archived.

Built on
Cloudflare WorkersD1R2Durable ObjectsSearch Console APIGA4MCP
Analyse my businessWhatsApp