Research Engineer, Infrastructure

Job Summary

Building and scaling infrastructure for post-training efforts, the full-time Research Engineer, Infrastructure will manage distributed training systems and internal tools in a remote environment.

Key responsibilities

• Scale the post-training stack to support SFT and RL on leading open-weight models
• Build distributed rollout infrastructure for model inference and training experience collection
• Own the performance and reliability of training runs, ensuring efficient GPU utilization and failure recovery

Required qualifications

• Experience in building ML infrastructure or distributed systems with a focus on performance and reliability
• Strong Python and software engineering skills, with familiarity in ML frameworks like PyTorch or JAX
• Experience with distributed training or inference and diagnosing performance bottlenecks
• Understanding of post-training workflows and experience with RL rollouts or agent environments is a plus
• Practical judgment in deciding what to build versus adopt, with the ability to collaborate with researchers