Machine Learning Systems Engineer, RL Engineering

Design, implement, and maintain the algorithms and infrastructure powering RL-based model finetuning, including RLHF workflows.
Improve speed, reliability, and usability of reinforcement learning training pipelines used by researchers.
Profile and optimize the end-to-end RL training workflow, diagnose slowdowns, and reduce bottlenecks.
Build tooling to launch training jobs in test environments and detect issues quickly.
Extend finetuning systems for new model architectures and enhance instrumentation to monitor performance, resource usage, and Python GIL contention.
Collaborate closely with researchers and engineers, contribute to a culture of safety and societal impact in AI systems, and practice pair programming.