Research Engineer - LLM Training Infrastructure - Seed Infra

Work on the Seed Infrastructures team to build and improve large-scale LLM training infrastructure for AI foundation models. Design distributed training strategies (parallelism, throughput scaling, and compute/communication optimizations) and improve reliability with fast checkpointing, fault tolerance, and failure diagnosis. Optimize network, scheduling, and GPU memory management across the training stack, analyze exascale bottlenecks, and translate research into production-ready AI infrastructure.