Principal ML Performance Engineer (GPU Optimization)

Proxima is a frontier AI and data generation company developing programmable protein interactions and proximity therapeutics through foundation-model machine learning and structural interactomics. The Principal ML Performance Engineer will profile and optimize structural and generative model training and inference, develop GPU kernels and profiling tools, scale distributed training, and improve GPU cluster efficiency and cost.

Responsibilities:

• Profile and optimize training and inference for structural and generative models, including transformers, diffusion, and geometric deep learning
• Write and tune custom kernels (CUDA, Triton) and use compilers (torch.compile, TensorRT, XLA) when beneficial
• Scale distributed training across 32-64 nodes, employing FSDP, DeepSpeed, tensor and pipeline parallelism, and mixed precision
• Reduce inference cost by optimizing memory scaling for large complexes, improving diffusion sampling efficiency, batching ragged inputs, and maximizing throughput across up to 1000 GPUs
• Manage GPU cluster efficiency on GCP, focusing on scheduling, utilization, spot strategy, and cost reporting
• Develop benchmarks and profiling tools for the research team