AIRemote (Global)Full-time
ML Engineer - Voice AI
Build and deploy ML models for answering machine detection (AMD), sentiment analysis, call scoring, and real-time voice transcription. You will work at the intersection of speech processing and production engineering.
Responsibilities
- Train and deploy AMD models with p99 latency under 3500ms
- Build real-time transcription and sentiment analysis pipelines
- Develop call quality scoring models using MOS and jitter metrics
- Optimize model inference for edge deployment on Supabase Functions
- Collaborate with the platform team to integrate ML into the dialer pipeline
Requirements
- 3+ years ML engineering with production model deployment
- Experience with speech/audio processing (ASR, TTS, speaker diarization)
- Proficiency in Python and familiarity with TypeScript
- Knowledge of model serving (ONNX, TensorRT, or similar)
- Published research or significant open-source contributions are a plus