AI Systems, Model Optimization

This is a fully remote position, open to applicants in California.

📋 Description

• Create the pathway from model architecture to physical silicon.

• Design the training methodologies, optimization techniques, and infrastructure necessary for efficient AI model execution on innovative compute platforms.

• Construct detailed performance models to assess trade-offs in compute, memory, and energy usage.

• Lead the partitioning and mapping of intricate AI models onto hardware.

• Develop and implement Quantization-Aware Training (QAT), noise-aware training, and sparsification methods.

• Enhance and optimize kernels utilizing low-level programming frameworks such as CUDA, Triton, or CUTLASS.

• Serve as a bridge between AI model architects and hardware/infrastructure engineering teams.

⛳️ Requirements

• An MS/PhD or equivalent research/project experience in a quantitative discipline, including AI/Machine Learning, Computer Science, Physics, Electrical Engineering, or Applied Mathematics.

• Profound practical knowledge of the contemporary AI/ML stack along with optimized compilation and execution of algorithms on current GPU systems.

• Demonstrated experience in profiling, identifying, and addressing performance bottlenecks in complex ML codebases.

• Proven ability to correlate cutting-edge AI model architectures (such as Transformers, Mixture of Experts, and diffusion models) with system performance outcomes.

• Extensive experience with PyTorch, including its internal workings, torch.compile, and distributed data parallel (DDP) / fully sharded data parallel (FSDP) libraries.

🏝️ Benefits

• A comprehensive package including best-in-class health benefits

• 401k matching

• Truly unlimited PTO

• Complimentary meals in our Palo Alto office