Director, Data Lead - Robotics and Physical AI

NVIDIA • Us, CA

14h

About The Position

NVIDIA has been transforming computer graphics, PC gaming, and accelerated computing for more than 25 years. It’s a unique legacy of innovation that’s fueled by great technology—and amazing people. Today, we’re tapping into the unlimited potential of AI to define the next era of computing. An era in which our GPU acts as the brains of computers, robots, and self-driving cars that can understand the world. Doing what’s never been done before takes vision, innovation, and the world’s best talent. As an NVIDIAN, you’ll be immersed in a diverse, supportive environment where everyone is inspired to do their best work. Come join the team and see how you can make a lasting impact on the world. NVIDIA is seeking a Data Lead to own the comprehensive data strategy for our Robotics and Physical AI efforts. As we move from generative intelligent systems in the digital world to Embodied AI in the physical world, data is the fuel that powers intelligence. Look beyond just one source, orchestrating a complete data engine that combines real-world human demonstrations, strategic data acquisitions, large-scale annotation, and physics-based simulation. The data at this stage involves sophisticated modalities—video, depth, proprioception, haptics, and natural language—used to train the next generation of Physical AI and World Foundation models. Dive into the heart of AI at NVIDIA. As a key player on our dynamic team, define how we teach robots to understand and interact with the world, balancing the fidelity of real-world data with the infinite scale of simulation!

Requirements

• Bachelor’s degree in Computer Science, Robotics, Engineering, or a similar technical field (or equivalent experience).
• 16+ overall years of experience in product management, data operations, or machine learning, specifically within hardware, robotics, or AV (Autonomous Vehicle) sectors.
• 8+ years of direct people management experience
• The "Full Stack" of Data: Experience managing diverse data pipelines—not just scraping the web, but dealing with hardware sensors, manual annotation services, and physical data collection logistics.
• Proven track record to manage external data vendors, annotation services, and strategic partners.
• Understanding of robotic sensors (LiDAR, IMU, RGB-D) and the concept of "embodiment" (how an agent moves through space).
• Ability to quickly experiment with data subsets to test hypotheses and determine the most efficient path to model accuracy.
• Ability to translate sophisticated robotics research requirements into actionable data collection plans for operational teams.

Nice To Haves

• Legal/IP Knowledge: Familiarity with data licensing, copyright for 3D assets, and intellectual property frameworks in AI.
• Foundation Model Experience: Experience working with VLA (Vision-Language-Action) models, Multimodal LLMs, or large-scale video pre-training.
• Tools Proficiency: Exposure to data management platforms (e.g., scale.ai, Labelbox, Voxel51) and 3D tools (Blender, Omniverse, USD).

Responsibilities

• Holistic Data Strategy & Pipeline Architecture: Author the Physical AI Data Ingestion Blueprint, defining the end-to-end logic for how data flows from the real world into our models. You will define requirements for de-duplication, versioning, and data lineage, ensuring the architecture supports massive scale.
• Strategic Acquisition & Qualification: Manage the external ecosystem of data partners. This involves establishing rigorous Qualification and Acceptance Criteria and managing Licensing to ensure all assets meet legal and technical standards before ingestion.
• Human Demonstration & Teleoperation: Lead the strategy for acquiring and validating high-quality human demonstrations, defining workflows for VR/hardware teleoperation, motion capture, and "human-in-the-loop" data collection.
• Multimodal Data Curation: Supervise sophisticated multimodal data, including, but not limited to: Vision: RGB, Depth, Semantics, Point Clouds.. Action: Robot manipulation paths and haptics , and 3D Assets: Neural reconstructions and USD/MJCF scenes.
• Sim-to-Real & Introspection: Collaborate with research to interpret model edge cases and failure modes. You will use these insights to refine the selection logic for data collection. Leverage reasoning models to assist in agentic data inspection.
• Annotation Operations: Manage the complex pipeline of 3D/4D annotation, working with vendors to define labeling standards for semantic segmentation and 6-DoF pose estimation.
• Data Governance & Ethics: Lead the development of internal policies regarding data privacy, consent, and bias to ensure responsible robot training.

What This Job Offers

• Job Type: Full-time
• Career Level: Director