Machine Learning / Reinforcement Learning Infrastructure Engineer

Eka Robotics develops intelligent robots for the physical world and is expanding its robotics and machine learning R&D team. The Reinforcement/Machine Learning Infrastructure Engineer will design, implement, and maintain large-scale model training systems, developer tooling, and distributed training infrastructure for robot learning. The role collaborates closely with researchers to scale experiments from prototyping to production and improve training workflows.

Responsibilities:

• Own Training Infrastructure: Design, implement, and maintain robust systems for large-scale model training, including job orchestration, scheduling, checkpointing, and experiment tracking
• Developer Experience & Tooling: Build streamlined, intuitive abstractions for launching, monitoring, debugging, and reproducing experiments, minimizing friction and maximizing productivity for our research teams
• Scale Distributed Training: Work closely with researchers to reliably scale reinforcement learning and machine learning pipelines across compute clusters
• Resource Management: Ensure efficient allocation and utilization of cloud-based compute resources while building the foundational systems needed for future scaling
• Collaborate with Researchers: Partner with the research team to understand their needs, build infrastructure that supports cutting-edge methods, guide best practices for training at scale, and contribute to core JAX model and training code