Voice AI Engineer

About the Role

We are seeking an experienced Voice AI Engineer to own our self-hosted LiveKit production stack. Unlike typical agent-only roles, this position requires deep expertise in running the entire infrastructure: server management, SIP services, Egress workers, and Agents on Kubernetes.

Key Responsibilities:

• Own and maintain the full LiveKit production stack (server, SIP service, Egress, Agent workers).
• Harden and configure SIP trunks for real carriers and contact-center platforms.
• Manage stateful agent workers including capacity planning, scaling, graceful draining, and deployments during live calls.
• Optimize latency from end-of-utterance to first audio across all components.
• Build comprehensive observability solutions to make call failures explainable post-call.
• Collaborate on STT/TTS/LLM pipelines, turn-taking logic, and interruption handling.

Tech Stack & Requirements:

• Mandatory: 1-2+ years Python (asyncio), Self-hosted LiveKit experience, WebRTC/SIP/RTP knowledge, Docker/Kubernetes.
• Nice to Have: Experience with Telnyx/Twilio/Vonage trunks under load, CCaaS integrations (Genesys/NICE/Five9), and media servers like FreeSWITCH or Janus.

The Setup

This is a fully remote role operating at high intensity. We expect 50-60 hours per week with primarily IST working hours but availability for US client meetings.