Staff Engineer, Inference Optimizations

📋 Description

• Lead benchmarking and performance optimization for inference.
• Engineer memory, precision management, and multi-node GPU parallelism.
• Apply cutting-edge optimization techniques to Gen AI workloads.
• Advise on hardware/software stacks (CUDA, ROCm, TensorRT) and procurement.
• Mentor engineers through code reviews and design guidance.
• Partner with Product and TPMs to translate hardware limits into features.

🎯 Requirements

• 5+ years in HPC or AI infrastructure.
• Gen AI literacy across LLMs, VLMs, and LMMs.
• Optimization: attention layers and distributed GPU parallelization.
• Hardware fluency: NVIDIA/AMD GPUs, CUDA/ROCm.
• Open-source experience building or contributing to OSS projects.
• Systems design: low-level GPU programming and memory access patterns.

🎁 Benefits

• Career development and access to learning resources.
• Competitive benefits and flexible time off.
• Support for conferences, training, and education.
• Inclusive, equal-opportunity workplace.