Senior ML Engineer: World Models for Robotic Simulation

Auxo Talent; San Francisco (CA); On-site; USD 150,000 - 275,000; Full time

Job summary

Auxo Talent is hiring a Senior Machine Learning Engineer to build world models and training infrastructure for robot manipulation policies in San Francisco. You will lead end-to-end development of video/dynamics models, enabling controllable, physically faithful simulations that scale across customers and tasks.

Expect a hands-on role with strong ownership: optimize training across multi-GPU clusters, write CUDA code, and push the boundaries of 3D vision and generative modeling in a

Qualifications

• Strong coding in Python and PyTorch, or equivalent
• Experience training image or video generation models end to end
• Track record of running training at cluster scale
• Knowledge of 3D vision, multi-view geometry and scene reconstruction

Responsibilities

• Train video and dynamics models for manipulation policies
• Develop architectures and training methods to improve rollout quality
• Own training infrastructure: multi-GPU clusters, custom CUDA
• Design and improve data engine to support learning across tasks

Skills

Python

PyTorch

CUDA

Multi-GPU training

Tools

CUDA

Job description

Auxo Talent is hiring a Senior Machine Learning Engineer to build world models and training infrastructure for robot manipulation policies in San Francisco. You will lead end-to-end development of video/dynamics models, enabling controllable, physically faithful simulations that scale across customers and tasks.

Expect a hands-on role with strong ownership: optimize training across multi-GPU clusters, write CUDA code, and push the boundaries of 3D vision and generative modeling in a