AI Performance Engineer (Cloud AI Engineering), Sr | Staff | Sr. Staff

What you'll do

• Convert, optimize, and deploy models for inference; analyze and improve LLM, VLM, and diffusion model performance for throughput and latency.
• Map AI workloads to current and future hardware, and develop high-level kernels such as those in Triton.
• Investigate performance and stability issues, build performance-insight tools, and collaborate with customers and internal compiler, firmware, and platform teams.

What they ask for

• Hands-on experience building and optimizing language models, particularly with PyTorch and ONNX.
• Strong understanding of transformer architectures, attention mechanisms, and performance trade-offs.
• Experience with workload mapping strategies involving sharding or other parallelisms; strong Python programming skills.
• Understanding of computer architecture, ML accelerators, in-memory processing, and distributed systems.

Nice to have

• Background in neural network operators, linear algebra, and math libraries; understanding of machine learning compilers.
• Experience with accuracy convergence and evaluation, torch.compile, or torchDynamo; PhD in Computer Science, Computer Engineering, or Machine Learning.

Pay

$194,400.00 - $291,600.00 USD base salary

Tools

PyTorch, ONNX, Triton, torch.compile, torchDynamo

AI focus

Inference acceleration and performance optimization for language, vision-language, and diffusion models, including GenAI algorithms such as attention mechanisms and MoEs.