MVP AI Voice Polling Platform

I’m building the first-round MVP of an Australian telephone polling platform and need a developer who can take it from concept to a working cloud-hosted product. The core workflow is straightforward: the system automatically dials a list of phone numbers, asks scripted questions, recognises spoken answers with NLP-based voice recognition, and stores each response for later analysis. It must comfortably interpret multiple-choice replies (e.g. “Press or say one for…”), open-ended comments, and numeric inputs such as postcode or age.

Key capabilities I need implemented:
• Automated outbound call scheduling with retry logic and time-zone awareness.
• Real-time voice recognition powered by Natural Language Processing so the caller can respond naturally rather than keying in touch-tones.
• A lightweight analytics layer that aggregates results and exposes them through a simple web dashboard plus downloadable CSV/JSON exports.

I expect the stack to be Python-centric—FastAPI or Django for the service layer, a PostgreSQL or similar store for call and response data, and a telephony gateway such as Twilio, Plivo or Asterisk for call handling. If you have a preferred combination that still delivers those three core features, I’m open to it.

Deliverables at the end of this phase:
1. Deployed MVP on my cloud account (AWS, GCP or Azure—your call).
2. Source code in a private Git repo with clean commit history and README-level documentation.
3. Written test plan plus unit tests covering the call flow and NLP response parsing.
4. Short video or live demo walking through a complete polling session and the reporting dashboard.

I’ll supply the call scripts, sample contact lists and a basic Figma wireframe for the dashboard. You’ll architect, code, integrate, and deploy. Let’s build a solid foundation that we can iterate on quickly.