Prime Intellect
Research Engineer - RL Infrastructure
$350K
San Francisco +1 · Full-time · Posted Oct 19, 2026
- Location
- San Francisco
Remote - Visa sponsorship
- Visa sponsorship and relocation assistance for international candidates.
About the job
Build and optimize distributed systems infrastructure for large-scale reinforcement learning and model training, including performance improvements across compute, memory, networking, and runtime layers.
What you'll do
- Develop and improve Prime Intellect’s prime-rl framework and RL training stack, including async rollout and post-training systems.
- Optimize training efficiency and performance across kernels, communication, memory, scheduling, and distributed workloads.
- Collaborate with researchers and infrastructure engineers and contribute to open-source and internal training systems.
What they ask for
- Strong systems engineering experience in AI/ML infrastructure, especially large-scale model training or inference.
- Deep familiarity with PyTorch and distributed training frameworks or related tooling.
- Experience optimizing training performance and hands-on experience with data, tensor, and pipeline parallelism.
- Strong understanding of GPU architecture, profiling, and performance debugging.
Nice to have
- Experience writing or optimizing CUDA / Triton kernels, compiler or runtime optimization, or RL training infrastructure.
- Experience with multi-node GPU clusters, high-performance networking, or contributions to open-source ML systems.
Pay
$150-350k (cash compensation, including equity incentives; currency not stated)
Tools
PyTorch, PyTorch Distributed, DeepSpeed, FSDP, Megatron, vLLM, Ray, CUDA / Triton
AI focus
Develop infrastructure for large-scale RL, distributed model training, async rollouts, and post-training systems for frontier-scale models.
Model-written summary · the original posting is authoritative