Robotics and Agent Solution Architecture Intern - 2026

Overview

• We're seeking outstanding interns to participate in our AI and accelerated computing projects.
• As an AI4Science Solution Architecture Intern, you’ll collaborate with world-class experts, contribute to groundbreaking innovations, and help build the future of artificial intelligence and high-performance computing.
• This is an outstanding opportunity to gain hands-on experience while working on real-world projects that make a significant impact! What you'll be doing: Use your skills in programming, AI, and accelerated computing to build innovative tools and applications in areas such as robotics, agentic modeling and AI model inference across many domains.
• Conduct AI engineering work, assist in developing and optimizing AI models and tools using NVIDIA SDKs and frameworks.
• Collaborating with internal teams and external researchers.
• Explore brand new trends in AI and computing acceleration to contribute to research and technology transfer projects.
• Be available 3–4 days per week for at least 6 months.
• Positions are primarily based in Beijing, Shanghai, or Shenzhen.
• What we need to see: Enrolled in a Master’s or Ph.D.
• program in Computer Science, Electrical Engineering, Applied Mathematics, or a related field.
• Solid programming experience in at least one language (Python, C/C++, etc.) and familiarity with Linux development environments.
• Strong analytical and problem-solving skills.
• Effective communication and a collaborative approach when working with multi-functional teams.
• Ways to stand out from the crowd: Familiar with large model inference frameworks or multi-modality models, knowledge of model inference benchmark.
• Hands-on experience or theoretical knowledge in robotics manipulation, locomotion, simulation, closed-loop validataion and models such as VLA/VAM.
• Familiar with various agent framework is a plus and knowledge of Openclaw, skills.
• Familiar with NVIDIA libraries (e.g., NemoClaw, Isaac, BioNeMo, Dynamo, PhysicsNeMo) as well as published research or open-source contributions in relevant areas.