AI Engineer: Build Voice AI System

We're building a voice AI product and need an experienced engineer to design the architecture and build a complete, working voice AI system from scratch. It needs to be production-ready enough to demo to a real client. This is not a "follow my spec" project. You'll make the architecture decisions yourself. We want someone who has built voice AI or real-time AI systems before and knows how to structure one cleanly. What you'll build: An end-to-end voice AI pipeline covering inbound call handling, speech-to-text, LLM-based intent understanding, and text-to-speech response. On top of that, appointment booking integration with an external scheduling system via API, SMS confirmation on successful booking, call logging to a database, and basic error handling and fallback logic for failed calls. You should have: Proven experience building voice AI or real-time conversational AI systems. Strong API integration skills across webhooks, async workflows, and third-party systems. The ability to design system architecture independently, not just implement a spec. Familiarity with tools like Twilio, Retell AI, Vapi, Deepgram, ElevenLabs, or equivalents. And clean, production-quality code with proper error handling, not a fragile demo. This isn't for you if: You've only built text-based chatbots and never worked with voice or real-time systems. You need a detailed technical spec before you can start. Or you haven't shipped a working AI system before. Deliverables: Working voice AI receptionist, tested end-to-end with real calls Source code with basic documentation covering setup instructions and an architecture overview Short walkthrough via a live call demonstrating the system working