Senior ML Engineer: GPU Inference & Low-Precision Training

Nebius; Greater London; On-site; GBP 90,000 - 130,000; Full time

Benefits offered by this job

Competitive pay

Career growth

Flexibility and ownership

Collaborative culture

Impactful AI projects

International environment

Job summary

Nebius is building a high‑performance AI cloud platform enabling scalable inference and fine‑tuning across thousands of GPUs. You will contribute to optimizing throughput, latency and cost-per-token while advancing the state of the art in large language models.

The role focuses on research and production integrations, collaborating with global teams to push foundation models to hardware limits and drive measurable performance gains.

Qualifications

• Proven expertise in ML theory and transformer models.
• Strong Python coding skills and experience with PyTorch or TensorFlow.
• Experience profiling GPU workloads (Nsight, PyTorch profiler, etc.).
• Understanding of GPU memory hierarchy and compute/memory tradeoffs.
• Familiar with MHA, RoPE, KV-cache, Flash Attention and quantisation.
• Experience with modern DL frameworks and software engineering practices (CI/CD, tests).
• Strong communication and leadership abilities.

Skills

Machine learning theory

Transformer architectures

Python

GPU profiling

Deep learning frameworks

CI/CD

Version control

Leadership

English communication

Tools

Nsight

PyTorch profiler

CUDA

Job description

Nebius is building a high‑performance AI cloud platform enabling scalable inference and fine‑tuning across thousands of GPUs. You will contribute to optimizing throughput, latency and cost-per-token while advancing the state of the art in large language models.

The role focuses on research and production integrations, collaborating with global teams to push foundation models to hardware limits and drive measurable performance gains.