Senior Inference Optimization ML Engineer

Highlights

• Own inference performance end-to-end for foundation models in production.
• Close the gap between training and real-world robot deployment.
• Build generalist robots that handle complex environments and edge cases.

What they offer

• Equity

Ideal candidate

A high-ownership, measurement-driven problem solver who loves diagnosing bottlenecks and squeezing efficiency from complex systems to make robots faster in the real world.

Requirements

• 3+ years in inference optimization, ML systems, or related field
• Deep hands-on experience with modern ML stacks (PyTorch required; JAX a plus)
• Strong understanding of compute, memory bandwidth, and I/O bottlenecks in large model inference
• Experience with quantization, distillation, pruning, and model compilation
• Familiarity with inference serving frameworks (Triton, TensorRT, vLLM, TorchServe)
• Exceptional debugging and measurement ability to isolate bottlenecks and validate improvements
• High ownership mindset and comfort in fast-moving environment