Solution Architect Intern, Robotics RL - 2025

NVIDIA accelerates humanoid robots development with Isaac solution and GR00T blueprint. We’re now looking for a robotics reinforcement learning expert to join NVIDIA China Solution Architect team, to engage and support local robotics partners and customers. As a Solution Architect Intern, you’ll collaborate with China Solution Architect teams to address the customer issues on robotics policy training in simulation environment, as well as policy evaluation in simulation or real environments. Your expertise will help robotics customers to build a successful NVIDIA practice.What you’ll be doingSetup the latest RL workloads and algorithms (locomotion mobility and whole-body control, integration with manipulation VLA) in Isaac Lab.Analyze policy training efficiency and convergency. Identify the factors that impact the training efficiency. Verify ways to improve the training efficiency.Analyze GPU sizing for policy training. Verify GPU sizing with distributed training.Leverage existing benchmark frameworks to setup pipeline to evaluate RL policy in both simulation and real environments.Verify methods to mitigate the sim2sim gaps and sim2real gaps.What we need to seePursuing BS, MS, or PhD. experience with robotics learning, Linux, DL, MLStrong academic background in Computer or Electrical Engineering, Computer Science, or related degreeExperience in using RL tools: Isaac Lab, Isaac Gym, or OpenAI GymExperience in using robotics simulation tools: Isaac Sim or MujocoDeep knowhow and experience in robotics policy training for locomotion or manipulationStrong experience in robotics policy evaluation in simulation and real environmentStrong experience with Python, Container, and LinuxExcellent communication and planning skills, while being self-motivated with a focus on execution and qualityWays to stand out from the crowdStrong background and rich project experience in robotics simulation and robotics learningRich problem-solving experience in reinforcement learning with Isaac LabStrong background and rich project experience in policy deployment sim2sim and sim2real gaps analytics and mitigation results