LLM Training Systems Engineer: Distributed GPU Infra

Job Title

Hyphen Connect is seeking a highly skilled LLM Pre-training & Distributed Systems Engineer in San Francisco, California. This role involves orchestrating large-scale training runs across 1,000+ GPUs and optimizing distributed infrastructure for efficient machine learning processes. Ideal candidates should possess deep expertise in 3D parallelism and strong systems engineering skills, particularly in C++, CUDA, and Python. The position is vital for automating and ensuring reliable training outcomes.