Title: Web Developer — Voice + 3D (Paid Discovery First, Then Trial Build)

Overview I'm building Wang Fei — a voice-first digital being users meet on the web. Backend is live on GCP (Firebase Auth + REST voice API). Docs in a private GitHub repo (read access after hire). I hire in two paid milestones. Trial build only starts after I approve Milestone 1. ——— MILESTONE 1 — Discovery (fixed fee, no coding) Fee: $100–150 (your quote in application) Deliverables (1–2 page written report, 2–3 days after repo access): • What exists today vs legacy lucy.banya.tech • Smallest path to demo on NEW API: https://banya4-20.web.app • How you would implement: Firebase sign-in + hold-to-talk + one voice turn + minimal 3D scene (R3F/Three.js) • Proposed Milestone 2 scope, timeline, and fixed fee • Risks / what you would NOT build in v1 Read: • API docs (private repo) • https://banya4-20.web.app/test.html • UX reference only: https://lucy.banya.tech/app (old stack; product is Wang Fei) ——— MILESTONE 2 — Trial build (only after I approve M1) Target: 5–7 days, fixed fee per your M1 proposal Deliverables: • Sign-in → hold-to-talk → play audio (V2 API) • Simple 3D scene (café/park) + basic gestures (wave, smile, etc.) • Short README: how to run locally Product feel: like a relaxed date in a café — presence + voice, not a chat widget. ——— Required to apply 1) Links to work showing voice UX and/or 3D web (R3F/Three.js). No links = not considered. 2) Your M1 fixed fee + earliest start 3) GitHub username Optional in application: rough sense of your M2 trial range (build only). Remote OK. Banya AI Limited (Dubai). — Gu Yibo