Inference Engineer, Robotics

About this role

OpenAI seeks a GPU Inference Engineer to optimize model serving and inference performance for their robotics research platform. You'll collaborate with researchers to build efficient serving infrastructure and design inference-friendly models while working with cutting-edge hardware and software systems.

What you'll do

• Optimize model serving, inference performance, and system efficiency for robotics applications
• Drive kernel-level and data movement optimizations to improve throughput and reliability
• Partner with research and product teams to ensure models scale effectively
• Design and build critical serving infrastructure for robotics growth
• Assist researchers in developing inference-optimized model architectures
• Navigate ambiguity and drive complex technical initiatives to completion

What they're looking for

• Model performance optimization, especially inference layer
• Kernel-level systems and low-level performance tuning
• GPU optimization and data movement
• Model serving infrastructure design
• Multimodal AI workload scaling
• Systems engineering and debugging
• Technical leadership and initiative ownership
• Python or C++ (implied)

Benefits

• Hybrid work model (3 days per week in office)
• San Francisco location with relocation assistance
• Work on AGI-level robotics research
• Collaborate with world-class AI researchers
• Equal opportunity employer with comprehensive non-discrimination policy