Open-Source phone Assistant Stack

Budget: $250 - $750

I need a complete, production-ready open-source stack that lets restaurants—and eventually any retail business—run a voice AI assistant with performance on par with commercial offerings.

Core capabilities I must see working end-to-end: • Taking orders accurately, including item modifiers and special instructions • Answering customer questions about the menu, hours, location, and policies • Providing personalized recommendations based on order history or stated preferences

The assistant has to run smoothly pick phone calls.

Here’s what I expect to receive:

1. Architecture diagram and tech stack rationale
2. Containerised or otherwise easily deployable codebase with setup scripts
3. Custom language and intent models tuned for restaurant scenarios
4. Demo app (Android and/or iOS tablet) showcasing all three user flows above
5. Clear documentation for ongoing training, maintenance, and feature expansion
6. Dashboard to show all statistics and transcript of the call etc.

Acceptance criteria: sub-second response time on contemporary mobile hardware, ≥95 % intent recognition accuracy in a typical restaurant noise profile, and hassle-free local installation with no proprietary licence dependencies.

If this sounds within your wheelhouse, let’s talk timeline and milestones.