ML Optimization Engineer: Training & Inference for Robotics

Generalist AI; San Mateo, Somerville (CA, MA); On-site; USD 180,000 - 240,000; Full time

Job summary

Generalist AI is seeking a highly capable ML optimization engineer to advance both training and on-robot inference stacks. You will push step-function gains across GPUs and software paths, shaping how models operate in real-world robotics tasks.

You will explore transformer and diffusion model optimizations, from CUDA kernels to Python-level improvements, aiming for substantial throughput with minimized latency in production pipelines.

Qualifications

• Experience with training and inference optimization for transformer and diffusion models.
• Proficiency with CUDA kernels and GPU optimization.
• Strong knowledge of Python, PyTorch, and NumPy.
• Experience optimizing both hardware and software paths.

Responsibilities

• Own optimizations on training and on-robot inference stacks.
• Implement ML, hardware, and software changes that enable step-function gains.
• Optimize both inference and training stacks.

Skills

ML optimization

CUDA kernels

Python performance

Transformer architectures

Diffusion models

NVLink

Tools

CUDA

PyTorch

NumPy

Python

Job description

Generalist AI is seeking a highly capable ML optimization engineer to advance both training and on-robot inference stacks. You will push step-function gains across GPUs and software paths, shaping how models operate in real-world robotics tasks.

You will explore transformer and diffusion model optimizations, from CUDA kernels to Python-level improvements, aiming for substantial throughput with minimized latency in production pipelines.