Research Engineer - Distributed Training
Core
Build and optimize distributed training infrastructure for pre-training and large-scale RL workloads to support frontier-scale AI model development.
Role type
Senior IC distributed training systems engineer
Builds
Distributed training systems, RL training stacks, and open-source ML infrastructure
Domain
AI/ML infrastructure, distributed systems, high-performance computing
Deliverable
production ML models
Required skills
distributed training frameworks, GPU architecture, performance profiling, parallelization strategies, systems engineering, CUDA/Triton optimization, compiler/runtime optimization
Preferred skills
RL training infrastructure, multi-node GPU clusters, high-performance networking, open-source contributions
Technologies
PyTorch, DeepSpeed, FSDP, Megatron, vLLM, Ray, CUDA, Triton
Responsibilities
Build and optimize distributed training infrastructure; Improve end-to-end training efficiency across compute, memory, networking, and scheduling; Design and implement low-level performance optimizations; Work on distributed training systems spanning data, tensor, and pipeline parallelism; Shape the architecture of the RL training stack; Contribute to open-source libraries; Collaborate with researchers to translate bottlenecks into system improvements.
Seniority
Senior, hands-on IC