Machine Learning Engineer | Audio, Speech, Acoustics & Sound
Core
Develop streaming, multilingual speech recognition, voice interaction, and audio understanding models for far-field audio captured by single-camera microphones in security devices.
Role type
Senior IC machine learning engineer (audio/speech)
Builds
Real-time speech recognition, voice activity detection, noise suppression, and sound event detection models deployed on edge camera devices.
Domain
Consumer security hardware, audio signal processing, deep learning
Deliverable
production ML models
Required skills
Deep learning for audio, audio signal processing (STFT, filterbank, mel features), speech recognition (CTC, transducer, LLM-based), model optimization (quantization, pruning, distillation), Python, PyTorch, Linux, multi-GPU training
Preferred skills
Far-field audio challenges (CHiME, DNS), on-device deployment (ONNX, TFLite, NPU), voice agents (ASR/LLM/TTS cascades), audiovisual models
Technologies
Whisper, Parakeet, Qwen3-ASR, PyTorch, ONNX, TFLite, sherpa-onnx
Responsibilities
Develop and fine-tune streaming speech recognition models; implement voice interaction features like end-of-turn detection; build noise suppression and dereverberation pipelines; create sound event detection and audio captioning models; design and evaluate small, real-time models for edge deployment; build audio data pipelines with synthetic mixing and augmentation; collaborate with firmware engineers for model integration.
Seniority
Senior, hands-on IC
