Vocal ERLab

ICG Labs · internal lab · phase 0

The emergency room for your voice

Voice rehabilitation built with a real voice coach. This lab is where every component gets proven as a standalone demo — before the product exists. Click everything; that's what it's for.

Component demos

live = real mic/camera, runs on your device

AI triage nurse

live

Voice-first intake — speak your answers and get routed to a starting program, with hard stops for red flags

PRD F6 · Phase 2

Vocal health check

live

Hold an “ahhh” — get an estimate of your voice qualities and whether a professional should take a listen

PRD F15 · Parallel track

Voice analyzer

live

Live pitch, clarity and spectrum while you hum — measured against your own baseline, on your device

PRD F8 · Phase 2–3

Practice recorder

live

Record yourself doing an exercise, review, re-record, send to coach

PRD F7 · Phase 2

Lesson scanner

live

Drop a lesson recording — see its acoustic fingerprint and add the context that makes it useful, all on-device

PRD F13 · Parallel track

Avatar studio

live

Nina's capture booth: consent-gated guided recording for her voice clone, then her video avatar

PRD F14 · Phase 2

Exercise library

Browse by symptom × who-you-are — the skippable front door

PRD F1 · Phase 1

Exercise player

One exercise, step by step, with completion and daily progress

PRD F2 · Phase 1

Program builder

Drag-and-drop: how providers assemble programs from existing exercises

PRD F9 · Phase 3

Scope library

The anonymized laryngoscopy research pool — one case, three disciplines learning from it

PRD F12 · Phase 3

Check-in reminders

SMS/email preferences — with deliberately health-free message bodies

PRD F5 · Phase 1–2

Booking

Book Nina at her rates — plus the bench of certified pros

PRD F4 · Phase 1

Disclaimer gate

The consent moment: wellness not medicine, logged acceptance

PRD F3 · Phase 1

Brand explorer

Vocal ER lockups, the ambulance palette, and the mascot question

PRD — · Open question

Product brief →

Everything decided in the founding conversation — two tiers, features, constraints, open questions.

Dev strategy →

How we build without drift: the environment ladder, the PRD rule, phases, and lessons from the ICG Labs portfolio.

PRD →

The spec layer. No feature gets built until its section exists here.

Research →

State of the art: the voice-AI landscape, compliance, competitors, and the reimbursement path.

How this lab works

  1. Every risky part of the product gets built here first, as something you can actually click.
  2. Reactions to these demos shape the spec — nothing graduates into the real app until it earns it.
  3. All content is synthetic — real curriculum, real users, and real data come later, on the right infrastructure.