AI Training Infrastructure - Member of Technical Staff
Core
Build and scale distributed training infrastructure supporting large AI models across large GPU clusters.
Role type
Senior IC distributed systems engineer (AI training infrastructure)
Builds
Distributed training systems, GPU clusters, model pipelines, and tooling for AI development
Domain
Cloud infrastructure + AI/ML training
Deliverable
infrastructure
Required skills
distributed training systems, large-scale machine learning infrastructure, multi-node GPU training, production ML pipelines, complex distributed systems programming
Preferred skills
PyTorch Distributed, DeepSpeed, Megatron-LM, Ray, supervised fine-tuning (SFT), reinforcement learning from human feedback (RLHF), Kubernetes, containerized AI workloads
Technologies
PyTorch, DeepSpeed, Megatron-LM, Ray, Kubernetes
Responsibilities
Build and scale distributed training infrastructure; Design systems to increase training reliability and efficiency; Develop solutions for fault tolerance and checkpointing; Integrate AI models into production training pipelines; Diagnose and resolve issues impacting training throughput and stability; Build tools and automation to improve developer experience; Establish best practices for training infrastructure operations
Seniority
Senior, hands-on IC