Software Engineer, RL Training Infra

Maintain and scale large RL training runs, debugging and improving training infrastructure.

• •Work to keep large-scale reinforcement learning (RL) training runs fast, reliable, and unblocked by debugging and improving training, inference, orchestration, and distributed systems.
• •Key Responsibilities Maintain and scale frontier RL training runs, addressing urgent engineering and infra problems.
• •Debug across training systems, inference, orchestration, and distributed infrastructure.
• •Improve reliability, efficiency, and tooling for RL training.
• •Support research-heavy integrations like multi-agent capabilities and memory.
• •Requirements Strong generalist engineering skills with ML infrastructure experience.
• •Experience debugging distributed systems and scaling model training.
• •High attention to detail, rapid learning, and strong communication.
• •Ability to operate under tight timelines and ownership of operational issues.