About the role We're hiring a founding engineer to build a production voice-AI customer service platform from the ground up, serving enterprise clients in an operations-heavy industry. This is a high-ownership, high-velocity role. You'll own the entire system, the real-time voice agent, the backend platform, the enterprise integrations, and the handoff into a live human customer service operation and you'll ship fast, working closely with the founder. What you'll build A production real-time voice agent: a streaming STT → LLM → TTS pipeline with reliable turn detection, interruption handling, and sub-second response latency. The full platform around it: backend services, an admin and analytics interface, and an event-driven architecture for call logging, transcripts, and QA. Tool/function calling that connects the agent to backend systems and enterprise CRMs, plus clean escalation and warm-transfer flows from the AI agent to human agents with full context handoff. An evaluation and iteration loop; testing the agent against real call patterns and continuously tuning behaviour. Requirements Strong full-stack engineering across a modern Type Script stack; able to design and run production services end to end. Hands-on experience building real-time voice or conversational AI — orchestration frameworks (e.g. Live Kit Agents), STT/TTS providers, telephony (SIP/Twilio), turn detection, and latency optimisation. Deep, practical command of LLM application engineering: prompt design and system-prompt architecture, tool/function calling, structured outputs, context management, retrieval, and evaluation. You understand how to make these systems reliable, not just demo-able. Fluent with agentic coding tools and AI-assisted development — you use them to ship at a pace well beyond a traditional engineer, and that output is the expectation, not a bonus. A track record of shipping integrations against complex enterprise and third-party APIs. Sound product judgment — you engage stakeholders directly, scope, and prioritise without dedicated product support. Comfort as the sole engineer: pragmatic, fast-moving, owns ambiguity.
Nice to have Domain exposure in logistics, field services, or other operations-heavy industries. Experience with contact-centre / BPO systems (CCaa S, IVR, agent-assist). Multilingual product experience.