Research Engineer (RL Infrastructure) Jobs in USA 2026

Ready to build the systems behind frontier-scale reinforcement learning? Prime Intellect seeks a Research Engineer for RL Infrastructure, in San Francisco or remote. Visa sponsorship and relocation assistance are offered for international candidates. If optimising distributed training and GPU performance excites you, this role offers real technical depth.

About the Role

You will build and optimise infrastructure for large-scale RL and distributed model training. This includes contributing to the open-source prime-rl framework. Additionally, you will improve efficiency across compute, memory, networking, and scheduling layers. You will also write low-level optimisations, including kernels, communication paths, and runtime improvements.

The work covers data, tensor, and pipeline-parallel training, plus asynchronous rollout systems. Cash compensation is listed at $150,000 to $350,000, including equity incentives.

About the Hiring Firm

Prime Intellect builds what it calls the open superintelligence stack, including its Lab platform. It was co-founded in 2024 by CEO Vincent Weisser and CTO Johannes Hagemann. In July 2026, it announced a $130 million Series A led by Radical Ventures. The posting says total funding has reached $150 million, from investors including Founders Fund and NVIDIA.

In November 2025, it released INTELLECT-3, an open 106-billion-parameter model trained with large-scale RL. Its Environments Hub hosts more than 2,500 community-built RL environments.

Job Duties

• Build and optimise infrastructure for large-scale RL and distributed training
• Contribute to the prime-rl framework and open-source libraries
• Improve training efficiency across compute, memory, and networking
• Implement kernel, communication, and runtime performance optimisations
• Help shape the architecture of the async rollout and post-training stack
• Work with researchers to turn bottlenecks into systems improvements

Job Requirements

• Strong systems engineering experience in AI/ML infrastructure
• Deep familiarity with PyTorch and distributed training tools such as DeepSpeed, FSDP, or Ray
• Experience optimising kernels, memory movement, or communication overhead
• Hands-on experience with data, tensor, and pipeline parallelism
• Strong understanding of GPU architecture, profiling, and debugging
• CUDA or Triton kernel experience is a plus

#339966;">#339966;" href="https://builtin.com/job/research-engineer-rl-infrastructure/11744408" target="_blank" rel="noopener">Click Here to Apply

Conclusion

This role suits systems engineers who love performance work at frontier scale. With open-source impact and strong backing, the work reaches far. For ML infrastructure engineers seeking a high-ownership role, Prime Intellect offers a compelling challenge.

#008000; text-decoration: underline;">READ MORE