Website:
neurodrift.ai
Job details:
Voice AI Engineer (LiveKit / Pipecat)
NeuroDrift is a US-based AI voice and enterprise software company. We build real-time voice AI agents that run live on real phone lines, at scale, for enterprise contact centres. Our platform carries production call traffic every day, so latency, telephony quirks and audio edge cases are our daily reality, not a research problem.
We are hiring an engineer who has built and run voice agents on LiveKit or Pipecat. Either framework counts.
Who this is for
If your voice AI experience is calling a hosted API or finishing a quickstart, this is not the role. We need someone who has:
- Built voice agents on LiveKit Agents or Pipecat and deployed them beyond a local demo
- Wired STT, LLM and TTS into a live pipeline and tuned how it behaves on real calls
What you'll do
- Build and ship voice agents end to end: audio pipeline, STT/TTS, LLM integration, tool calling
- Run and improve our self-hosted voice stack in production: media server, SIP, agent workers on Kubernetes
- Measure and reduce latency from end-of-utterance to first audio
- Tune turn-taking: endpointing, VAD, barge-in and interruption behaviour
- Debug bad calls across telephony, media and agent logs
What we need
- 1.5+ years Python, including async (asyncio, FastAPI or similar)
- LiveKit Agents or Pipecat in a real deployment
- STT, TTS and LLM providers in a live voice pipeline
- WebRTC, SIP, RTP fundamentals
- Docker; debugging from logs and metrics
Nice to have
- Self-hosted LiveKit server on Kubernetes: config, Redis, TURN/ICE, upgrades
- SIP trunk integration: LiveKit SIP, Twilio, Telnyx or similar
- Speech vendor tuning: Deepgram, Whisper, Azure Speech, keyterm biasing, telephony-band audio
- Contact-centre integration: Genesys, Five9, NICE, SIP REFER, transfers
- Latency instrumentation and per-minute STT/TTS cost awareness
How we work
- Claude Code from day one
- Proof of concept first, then hardening on real phone calls
The setup
- Fully remote
- High intensity, 50 to 60 hours a week. Startup pace, not a 9-to-5
- Working hours primarily IST, with availability for US client meetings
- Immediate joiners
This role is for you if you have built voice agents that real people have talked to, and you want to make them faster and more human.
Click on Apply to know more.