Senior Machine Learning Engineer, Model Training and Reinforcement Learning

nebius · Palo Alto, California, United States · Palo Alto, California, United States

<div class="content-intro">About Nebius: Nebius is leading a new era in cloud infrastructure for the global AI economy. We are building a full-stack AI cloud platform that supports developers and enterprises from data and model training through to production deployment, without the cost and complexity of building large in-house AI/ML infrastructure. Built by engineers, for engineers. From large-scale GPU orchestration to inference optimization, we own the hard problems across compute, storage, networking and applied AI. Listed on Nasdaq (NBIS) and headquartered in Amsterdam, we have a global footprint with R&D hubs across Europe, the UK, North America and Israel. Our team of 1,500+ includes hundreds of engineers with deep expertise across hardware, software and AI R&D.</div><h3>The role </h3> Nebius Token Factory is building an AI training and model post-training capability for frontier model improvement. This role owns the infrastructure that makes large-scale training and RL experiments possible, reliable, reproducible, and efficient. The work sits at the intersection of distributed systems, GPU performance, model training frameworks, RL pipelines, and production engineering. A Senior Machine Learning Engineer owns substantial ML work end to end. They can translate an ambiguous capability goal into concrete experiments, implement and debug training and RL recipes, build the supporting data and systems, and deliver measurable improvements in model quality, experiment throughput, and reliability. They are deeply hands-on and can independently debug both model-behavior failures and distributed training failures. Your responsibilities:  <ul class="ak-ul" data-local-id="11d761bd-5698-433f-9b21-a7fee44fb25c" data-indent-level="1"> <li> Design and run model-training and post-training experiments, including SFT, continued pretraining, preference optimization (DPO/IPO/KTO), and RL methods such as RLHF/RLAIF, PPO, and GRPO. </li> <li> Build reward functions, judge models, verifiers, task environments, and evaluation sets for reasoning, coding, tool use, and agentic workflows. </li> <li> Create synthetic data and data pipelines, including teacher-student generation, self-play, rejection sampling, filtering, and quality scoring. </li> <li> Analyze model-behavior failures and turn them into targeted data, reward, or algorithm improvements. </li> <li> Build and maintain distributed training and RL infrastructure using frameworks such as Megatron-LM, DeepSpeed, PyTorch FSDP/DTensor, Ray, verl, slime, AReaL, or OpenRLHF. </li> <li> Implement and debug parallelism strategies (tensor, pipeline, sequence/context, expert, and data parallelism) and build reliable rollout, reward-serving, checkpointing, and experiment-orchestration components. </li> <li> Profile and improve GPU utilization, memory usage, communication efficiency, training throughput, and inference/serving performance. </li> <li> Design rigorous evaluations and ablations for capability, instruction following, reasoning, tool use, safety, and regression risk. </li> <li> Write clear experiment plans, design docs, benchmark reports, and runbooks, and partner across research and platform teams. </li> </ul> Must-haves:

Let Nod apply for you →

More jobs at nebius

Sr. Enterprise Network Engineer — nebius Senior Technical Project Manager (Region Delivery) — nebius Site Reliability Engineer — nebius Startup Program Lead — nebius System Engineer (Compute Node) — nebius Technical Program Manager (Post-Sales) — nebius