Research Engineer - RL Infrastructure

Role overview Build and optimize the systems that make large-scale reinforcement learning and distributed model training practical. The work spans hardware-aware performance, communication, runtime, and scheduling, with contributions to an open training framework used for frontier-scale workloads.

Responsibilities - Develop infrastructure for large-scale RL and distributed training. - Improve end-to-end training efficiency across compute, memory, networking, and scheduling. - Implement low-level performance improvements in kernels, communication paths, and runtimes. - Contribute to distributed training systems and the associated open-source framework. - Investigate system bottlenecks and translate findings into measurable improvements.

Requirements - Strong systems engineering and performance optimization skills. - Experience with distributed systems or large-scale machine learning workloads. - Ability to work across compute, memory, networking, and runtime layers. - Familiarity with low-level optimization, such as kernels or communication primitives. - Interest in building and maintaining open-source infrastructure.

Benefits and work setup - Full-time role based in San Francisco or remote. - Visa sponsorship and relocation support are available for international candidates. - Quarterly team gatherings, hackathons, conference participation, and learning opportunities.