Research Engineer - Inference

Job Summary

Building inference systems that power cutting-edge RL research, the full-time Research Engineer - Inference will optimize GPU inference systems and improve rollout throughput for large-scale models, with the flexibility to work remotely or onsite in San Francisco.

Key responsibilities

• Extend vLLM integration for new model architectures and validate on small scales before larger runs
• Optimize rollout throughput for long-horizon agents by tuning concurrency and KV-cache reuse
• Collaborate with model researchers to enhance inference efficiency while balancing model quality and execution cost

Required qualifications

• Experience building or optimizing GPU inference systems and understanding inference-engine internals
• Strong Python skills with experience in PyTorch or a comparable ML framework
• Ability to profile end-to-end workloads and implement verified performance improvements
• Experience debugging numerical or integration issues in ML systems