Research Engineer - Speech/Audio Machine Learning
Core
Bridge research and production constraints to deliver competitive performance for speech/audio machine learning systems, handling end-to-end performance from scalable training infrastructure to low-latency inference.
Role type
Research Engineer (Speech/Audio ML)
Builds
Scalable training infrastructure, optimized model graphs, data pipelines, and high-performance demo environments for real-time audio interactions.
Domain
AI / Deep Learning / Speech & Audio
Deliverable
production ML models
Required skills
Systems Programming, Computer Architecture, Model Optimization (TensorRT, Apache TVM, OpenVINO), Model Compression (Quantization, Pruning), Distributed Training (DeepSpeed, Horovod), Data Pipeline Engineering, Containerization (Docker), WebRTC integration
Preferred skills
null
Technologies
TensorRT, Apache TVM, OpenVINO, DeepSpeed, Horovod, Docker, WebRTC
Responsibilities
Optimize model compute graphs for target runtimes; Perform quantization and pruning for real-time execution; Build reliable data pipelines for audio datasets; Design efficient multi-node training scripts; Build and maintain high-performance demo environments with real-time audio protocols.
Seniority
Mid-level, hands-on IC