ProjectGina
prototype · all data is fixture

Demo map — a routing platform for healing work

The premise: the client never browses, compares, or chooses a practitioner. One diagnostic read, a human placement, a record that travels, and a sub-60-second post after every session. Walk the tour in order — it follows one client, M., through the system.

The tour · ~10 minutes
01 Public site What a prospective client sees. Notice: no directory, the guarantee up front, published numbers instead of trust badges. 02 The door — walk in Same-day entry with a three-question triage gate. Answer "yes" to any question to see the 988 handoff; the door is honest about staffed hours. 03 Client portal — M. "Why Linder, why now," tokens that travel, the readable shared record, pre-session check-in, checkpoint request, Stripe-style purchase. 04 Practitioner portal Linder's view: upcoming placements, coverage, filing record, her terms. Open the brief from here. 05 The two-stage brief Facts and the working note first. Prior impressions stay locked until she logs her own read — try the gate. 06 The post — time it The sub-60-second filing that feeds the next brief. A live timer runs — the whole routing intelligence depends on this clearing 60 seconds. 07 Operator console Reads queue, routing signals, R.T.'s placement (hard exclusions vs preferences, the "why" the client hears verbatim), the network bench. 08 The read form What the operator fills during the call. Tick hard exclusions and watch the eligible pool shrink — it warns below three. 09 Metrics & kill conditions Cycle time as the headline, unit economics with live CAC entry, and all seven kill conditions published red or green. 10 Full scope — Tracy The operator's printable writeup that travels with every placement in a client's arc.
Working in this prototype
  • Every interaction on every screen — gates, timers, toggles, purchases, triage
  • The post form's 60-second timer, for live testing with practitioners
  • Walk-in triage routing and staffed-hours logic (Mountain Time)
  • Record export, CAC-to-margin math, eligible-pool math
Deliberately not built yet
  • Accounts, database, payments — all data is one hard-coded client arc
  • The AI brief compiler — sample text stands in for generated output
  • Placement automation — manual by design, per the operating model
  • Scheduling, messaging, notifications