All jobs
Staff Machine Learning Engineer, LLM Inference Optimization
Nebius
Zurich +4 · On-siteSenior (5–10 years)Full-timePosted 15 days ago
About the job
What you'll do
You will own the end-to-end path from model artifacts to fast, reliable, and cost-efficient inference endpoints. This involves optimizing LLM and VLM performance through techniques like quantization, speculative decoding, and engine selection while providing technical leadership to cross-functional teams.
What they're looking for
The role requires strong Python and PyTorch engineering skills with hands-on experience in deploying or optimizing high-throughput transformer inference systems. Candidates must have a deep understanding of inference bottlenecks and experience with modern inference stacks like vLLM or TensorRT-LLM.
Full description