
A service professional cannot assess a vague quote request without asking what work is needed, which equipment is involved, where the job is, what access is available and when the caller prefers. QuoteReady gathers those five details through an AssemblyAI voice conversation and prepares a draft the caller can inspect. Each answer retains supporting transcript evidence. Unknown answers and refusals stay explicit. A correction invalidates the previous review, and the caller must approve the current version before downloading a JSON request. The prototype does not price, book, email or promise a callback. The original JavaScript/Node.js app uses temporary server-issued tokens, WebSocket audio, Web Audio resampling and client-side function tools. The public live app is protected by a private reviewer invitation code and three-minute session limits. An open interactive sample and captioned demonstration require no code. On September 8, the public browser completed all five fields using synthesized fictional speech, corrected Friday afternoon to Monday morning and downloaded the reviewed draft. The 64-second recording shows actual AssemblyAI replies, transcripts and app state through the corrected readback; approval/download was verified separately. A basic microphone connection also transcribed input and returned a reply. Broader real-speaker and interruption evaluation is still future work. All 26 affected local tests passed. Next steps are evaluation with service businesses and optional business-side delivery with explicit caller approval. No customer adoption, revenue or time savings are claimed.
30 Sep 2026