Research Scientist / Engineer – Reinforcement Learning Infrastructure

Highlights

• Build the systems that make reinforcement learning work at frontier scale.
• Design agentic environments for multi-step tasks scaling to millions of episodes.
• Keep large RL runs stable with evaluation, monitoring and debugging tooling.

Ideal candidate

A systems-minded engineer who has lived through large-scale RL post-training and loves making fast, stable, correct distributed loops. Thrives debugging async pipelines and turning research ideas into production runs.

Requirements

• Hands-on experience post-training LLMs with RL (PPO/GRPO-family, RLHF, RLVR) at meaningful scale
• Extensive distributed PyTorch training and parallelism (FSDP, Tensor/Pipeline/Expert Parallel) for foundation models
• Experience building RL environments, reward functions, verifiers, or evaluation harnesses for LLM agents, including sandboxed execution and multi-turn tool use
• Deep familiarity with RL post-training frameworks (veRL, OpenRLHF, TRL, Ray orchestration) and rollout inference engines (vLLM, SGLang)
• Strong understanding of GPU clusters, networking, and communication libraries (NCCL, MPI) under mixed training and inference workloads