On-Device Transformers Tech Lead for Edge Inference

OpenAI; Los Angeles (CA); Hybrid; USD 400,500 - 489,500; Full time

Job summary

A leading AI research organization in San Francisco seeks an Inference Technical Lead to evaluate silicon platforms and work on model deployment for edge devices. You will collaborate with top machine learning researchers to push the boundaries of model capabilities. This role requires experience with GPUs and NPUs, understanding transformer models, and leading teams on performance-critical software. Competitive compensation package, including equity, is offered, along with a hybrid work model.

Qualifications

• Experience evaluating or deploying workloads on GPUs, NPUs, or specialized accelerators.
• Understand performance characteristics of transformer models including attention and memory bandwidth.
• Design or optimize high-performance compute systems like inference engines and distributed runtimes.

Responsibilities

• Evaluate and select silicon platforms for on-device deployment of models.
• Work closely with research teams to co-design model architectures.
• Analyze system performance tradeoffs between design and hardware capabilities.
• Lead a team responsible for implementing the low-level inference stack.

Skills

Experience with GPUs and NPUs

Understanding of transformer model performance

Designing high-performance compute systems

Leading teams in performance-critical software

Job description

A leading AI research organization in San Francisco seeks an Inference Technical Lead to evaluate silicon platforms and work on model deployment for edge devices. You will collaborate with top machine learning researchers to push the boundaries of model capabilities. This role requires experience with GPUs and NPUs, understanding transformer models, and leading teams on performance-critical software. Competitive compensation package, including equity, is offered, along with a hybrid work model.