Remote RL Infrastructure Engineer for Frontier AI Systems

Prime Intellect, Inc.; Oregon; Hybrid; USD 150,000 - 350,000; Full time

Benefits offered by this job

Cash compensation range

Remote or SF office work

Visa sponsorship and relocation

Quarterly off-sites & learning

Team learning opportunities

Job summary

Prime Intellect, Inc. is building the open frontier AI stack, unifying compute, environments, and secure sandboxes for frontier-scale model training and deployment.

The role focuses on scaling RL training systems, optimizing performance across kernels, memory, and networking, and shaping the architecture of the training stack with researchers and engineers. You will contribute to open-source libraries and internal infrastructure, collaborating closely with researchers to translate bottlenecks

Qualifications

• Experience with large-scale model training or inference.
• Strong systems engineering mindset for AI infrastructure.
• Deep knowledge of GPU architectures and performance tuning.

Responsibilities

• Build and optimize RL/distributed training infrastructure.
• Improve end-to-end training efficiency across compute, memory, networking, and scheduling.
• Implement low-level performance optimizations and runtime improvements.
• Work on distributed training systems spanning data, tensor, and pipeline parallel workloads.
• Shape architecture of the RL training stack and post-training systems.
• Contribute to open-source libraries and internal infra used for frontier-scale training.

Skills

Systems engineering

PyTorch

Distributed training

GPU architecture

Performance debugging

Data parallelism

Tensor parallelism

Pipeline parallelism

Bottleneck analysis

Ownership in fast-moving env

Tools

PyTorch Distributed

DeepSpeed

FSDP

Megatron

vLLM

Ray

CUDA

Job description

Prime Intellect, Inc. is building the open frontier AI stack, unifying compute, environments, and secure sandboxes for frontier-scale model training and deployment.

The role focuses on scaling RL training systems, optimizing performance across kernels, memory, and networking, and shaping the architecture of the training stack with researchers and engineers. You will contribute to open-source libraries and internal infrastructure, collaborating closely with researchers to translate bottlenecks