Senior ML Engineer, Inference & Optimization

Nebius Group; Zürich; Vor Ort; CHF 150.000 - 220.000; Vollzeit

Benefits dieser Stelle

Competitive compensation

Career growth and learning机会

Flexibility and ownership

Collaborative and innovative culture

Opportunity to work on impactful AI项目

International environment and talented

Zusammenfassung

Nebius Group in Zürich seeks a Senior Machine Learning Engineer to own model and endpoint optimization from artifacts through production deployment. You will work on model internals, inference engines, serving architecture, and benchmarking to improve latency, throughput, memory efficiency, GPU utilization and cost per token while maintaining model quality.

This is a hands‑on role that collaborates with kernel and platform engineers to diagnose bottlenecks, compare serving configurations, and

Qualifikationen

• Strong Python and PyTorch engineering skills.
• Hands-on experience deploying or optimizing LLM/VLM systems.
• Knowledge of modern inference stacks such as vLLM, TensorRT-LLM, Triton.

Aufgaben

• Own optimization work for model families, endpoints, or serving backends.
• Run engine comparisons and recommend serving configurations.
• Debug quality or performance regressions during production rollouts.
• Deploy, configure, benchmark, and extend inference engines (e.g., vLLM, Triton).
• Build reproducible benchmarks for latency, throughput, and cost per token.

Kenntnisse

Python

PyTorch

LLM deployment

Performance optimization

Inference systems

Tools

vLLM

SGLang

TensorRT-LLM

Triton Inference Server

NVIDIA Dynamo

Ray Serve

KServe

Jobbeschreibung

Nebius Group in Zürich seeks a Senior Machine Learning Engineer to own model and endpoint optimization from artifacts through production deployment. You will work on model internals, inference engines, serving architecture, and benchmarking to improve latency, throughput, memory efficiency, GPU utilization and cost per token while maintaining model quality.

This is a hands‑on role that collaborates with kernel and platform engineers to diagnose bottlenecks, compare serving configurations, and