Senior ML Engineer: Scalable Training & RL Systems

Nebius Group; Palo Alto (CA); Presencial; USD 195 000 - 262 000; Tempo integral

Vantagens oferecidas por esta oferta de emprego

Health insurance

401(k) plan with match

Parental leave

Remote work reimbursement

Disability & life insurance

Resumo da oferta

Nebius Group is seeking a Senior Machine Learning Engineer to own large-scale training and RL workflows for AI model development. You will design experiments, build data pipelines, and optimize distributed training infrastructure—collaborating closely with researchers and platform teams.

The role emphasizes hands-on model training, RL pipelines, and efficient GPU usage, with a focus on reliability and measurable improvements in model quality and throughput.

Qualificações

• Strong Python and PyTorch engineering skills.
• Experience in model training, post-training/RL, or large-scale ML systems.
• Ability to design rigorous experiments with baselines, ablations, metrics.
• Practical understanding of LLM behavior, instruction tuning, and evaluation challenges.
• Understanding of transformer training bottlenecks, memory management, and checkpointing.
• Ability to reason quantitatively about model quality, throughput, reliability, and cost.
• Strong communication and collaboration with researchers, engineers, and leadership.

Responsabilidades

• Design and run model-training and post-training experiments (SFT, continued pretraining, RL methods).
• Build data pipelines and evaluation sets for reasoning, coding, tool use, and agentic workflows.
• Analyze model-behavior failures and translate into data/algorithm improvements.
• Maintain distributed training and RL infrastructure using Megatron-LM, DeepSpeed, Ray, or OpenRLHF equivalents.
• Implement and debug parallelism strategies and orchestration components.
• Profile and improve GPU utilization, memory, and training throughput.
• Write experiment plans, design docs, and benchmark reports; collaborate with research and platform teams.

Conhecimentos

Python

PyTorch

Deep learning

Distributed systems

Model training

RL pipelines

Experiment design

Communication

Ferramentas

Megatron-LM

DeepSpeed

PyTorch FSDP/DTensor

Ray

Slurm

Kubernetes

Descrição da oferta de emprego

Nebius Group is seeking a Senior Machine Learning Engineer to own large-scale training and RL workflows for AI model development. You will design experiments, build data pipelines, and optimize distributed training infrastructure—collaborating closely with researchers and platform teams.

The role emphasizes hands-on model training, RL pipelines, and efficient GPU usage, with a focus on reliability and measurable improvements in model quality and throughput.