Senior RL Systems Engineer - Scaled Infrastructure

Speedrun Talent Network; Union (NJ); On-site; USD 180,000 - 240,000; Full time

Job summary

Luma is seeking an experienced Senior RL Systems Engineer to design, build, and scale distributed RL post-training systems across thousands of GPUs. You will own rollout generation, environment design, reward infra, and tooling to monitor and debug large RL runs.

The role emphasizes productionizing post-training ideas and collaborating with researchers to improve training efficiency. Applicants should have hands-on experience with RL at scale, distributed PyTorch, and RL frameworks, plus

Qualifications

• Hands-on experience post-training LLMs with RL at meaningful scale.
• Extensive distributed PyTorch training and parallelism for foundation models.
• Experience building RL environments, reward functions, verifiers, or evaluation harnesses for LLM agents.
• Familiarity with RL post-training frameworks and rollout inference engines.

Responsibilities

• Design, build, and scale distributed RL post-training systems across thousands of GPUs.
• Build high-throughput rollout generation, integrating inference engines and synchronization.
• Design RL environments for multi-step tasks with sandboxed execution and tool use.
• Build reward infrastructure and defenses against reward hacking.
• Develop evaluation, monitoring, and debugging tooling for large RL runs.
• Advance training efficiency and turn new ideas into production runs with researchers.

Skills

RL at scale experience

Distributed training experience

Post-training LLMs

Tooling and sandboxed execution

GPU cluster know-how

Tools

vLLM

SGLang

veRL

OpenRLHF

TRL

Ray

NCCL

MPI

Job description

Luma is seeking an experienced Senior RL Systems Engineer to design, build, and scale distributed RL post-training systems across thousands of GPUs. You will own rollout generation, environment design, reward infra, and tooling to monitor and debug large RL runs.

The role emphasizes productionizing post-training ideas and collaborating with researchers to improve training efficiency. Applicants should have hands-on experience with RL at scale, distributed PyTorch, and RL frameworks, plus