Agent RL Infra Engineer

What you'll do

• Evaluate and adapt emerging RL approaches into enterprise-ready blueprints.
• Design verifiable reward environments using NeMo Gym for internal use cases.
• Operationalize NVIDIA training backends as production services within Sandbox.
• Integrate with NeMo Microservices to enable end-to-end data flywheel workflows.
• Lead data curation strategies to continuously improve the quality of training data.

What we're looking for

• MS in CS, ML, or related field (or equivalent experience) and 10+ years of relevant industry experience
• Experience operationalizing RL techniques like DPO, GRPO, PPO, RLAIF into reusable workflows
• Proficiency with Python, Go, Rust, and familiarity with distributed training frameworks such as Megatron, NeMo, DeepSpeed
• Strong background in ML ops including pipeline automation, job orchestration, and GPU cluster management
• Experience building RL environments or training recipes for self-service consumption by other teams
• Familiarity with NVIDIA infrastructure (DGX, AI Factory) and NeMo Microservices preferred