Speech AI Engineer
Skills
Artificial IntelligenceGraphics Processing UnitHugging FaceMachine LearningPytorchPythonResearch Skills
What the job involves
The main requirements, responsibilities and hiring steps.
Requirements
- 5+ years in applied ML
- 2+ years working directly on speech models
- Strong Python and PyTorch skills
- Hands-on experience with Hugging Face and PEFT methods
- Working knowledge of CUDA and GPU performance optimization
- Production experience with vLLM or similar serving stacks
- Experience taking models from research to production
- Solid grasp of speech evaluation methodology
- Experience with multilingual or lower-resource European languages is a strong plus
- Experience with generative model architectures is a strong plus
- Experience with distillation pruning or speculative decoding is a strong plus
- Familiarity with telephony audio and codec constraints is a strong plus
- Production experience with Triton TensorRT or ONNX Runtime is a strong plus
- Distributed training experience is a strong plus
- Open-source contributions to speech or generative AI projects is a strong plus
- Exposure to RAG and agentic systems is a useful plus
Nice to have
- Senior IC
- Hands-on builder
- Research-minded
- Performance-focused
- Collaborative
- Entrepreneurial
Day to day
- Own end-to-end speech AI work across ASR and TTS for real-time voice products
- Fine-tune, evaluate, and deploy streaming speech models that meet strict latency and quality targets
- Optimize inference pipelines and GPU performance using modern serving stacks and low-level acceleration