Senior Machine Learning Engineer - Physical AI and Synthetic Data Generation and Evaluation

Details

We are looking for outstanding Machine Learning Engineers to join our Physical AI teams. This role develops sophisticated reasoning modules and high-fidelity synthetic datasets for AI agents that interact with the physical world, including technology supporting autonomous vehicles.

Responsibilities

• Develop advanced image and video generation, editing, and reasoning models for Physical AI applications.
• Build and fine-tune large-scale multimodal models, including VLMs, MLLMs, and generative models, using transformer, autoregressive, and diffusion-based architectures.
• Process visual and structured inputs, such as world model representations, and analyze consistency between user-intended scenarios and generated data.
• Apply and evolve user controls during data generation to provide precise environmental and structural control.
• Build and test automated quality assurance pipelines for sensor data and ego policies using MLLMs and classical algorithms.
• Develop capabilities to evaluate behavioral policies and ensure high-quality data delivery for VLA systems.
• Establish KPI evaluation and validation processes for the quality and physical accuracy of synthetic data releases.
• Create benchmark datasets and design and validate KPI metrics.
• Lead the generation of massive training datasets using state-of-the-art tools and synthetic data mining techniques.
• Contribute to the full lifecycle of machine learning software, including performance optimization, testing, and documentation.

Requirements

• BS, MS, or PhD in Computer Science, Computer Graphics, Robotics, or a related field, or equivalent experience.
• 12+ years of experience in machine learning software development.
• Deep technical knowledge of image and video synthesis, including diffusion models and state-of-the-art multimodal methods.
• Strong hands-on skills with major deep neural network libraries and programming languages, including Python.
• Experience with workflow management and databases for large-scale training and data generation.
• Strong code-efficiency optimization skills are a plus.
• Strong analytical and mathematical skills, including the ability to connect data-driven approaches with physical-world constraints.
• Experience assessing the impact of synthetic data on model performance through metrics and systematic validation.
• Collaborative approach and outstanding communication skills.

Preferred Qualifications

• Experience with computer or GPU architecture to improve inference and training performance.
• Familiarity with simulation platforms.
• Deep understanding of 3D sensor modalities, including cameras, multi-camera systems, lidar, and radar.
• Experience with open-source software.

Benefits

The role includes eligibility for equity and benefits.