Solutions Architect - AI Technology Center for Physical AI

Responsibilities

• Act as an elite research partner to catalyze the global research community by embedding NVIDIA as a core scientific collaborator.
• Drive the adoption of NVIDIA's platform through high-impact Physical AI engagements in Digital Twin and Robotics.
• Help researchers in universities and enterprises build virtual worlds using NVIDIA Omniverse (OpenUSD), generate synthetic data with NVIDIA Cosmos, and test robots using NVIDIA Isaac Sim.
• Advise on the transition from experimental code to large-scale, production-ready research environments.
• Review and rank research proposals based on technical soundness, novelty, and alignment with NVIDIA reference designs.
• Fine-tune and deploy NVIDIA GR00T models and other humanoid foundation models.

Requirements

• PhD in Computer Science, Electrical/Computer Engineering, Physics, Mathematics, or a related engineering field.
• 5+ years of hands-on experience in Digital Twin and Robotics technologies.
• Solid understanding of deep learning, including computer vision, LLMs, VLMs, or robot foundation models.
• Strong technical background in 3D virtual reality, photorealistic technologies, and USD (Universal Scene Description).
• Proficiency with ROS2, Isaac ROS, and NVIDIA Jetson/JetPack SDK.
• Experience working with NVIDIA GPUs and software libraries.
• Excellent presentation, collaboration, and written/oral communication skills in English.

Preferred Qualifications

• Proven track record of academic research publications.
• Experience working across academia, industry, and government sectors.
• Deep expertise in embedded systems, specifically the NVIDIA Jetson platform (ARM CPU, GPU, JetPack SDK).
• Experience with dexterous/bimanual manipulation, motion retargeting, teleoperation, and humanoid whole-body control.

About the Company

NVIDIA is a global leader in AI computing. The NVIDIA AI Technology Center seeks to drive frontier AI discoveries by partnering with the world's leading researchers and developers to solve the "three-computer problem" in Physical AI: simulating in the virtual world, training models, and deploying them into embedded systems.