Robotics Simulation & Synthetic Data Intern

Deccan AI•Mountain View, CA
•Onsite

About The Position

Deccan AI is hiring an intern for its Physical AI team to build synthetic data for embodied AI. The role owns the end-to-end synthetic data generation pipeline, from configuring simulation environments in NVIDIA Isaac Sim and Isaac Lab, through GR00T-Mimic trajectory expansion, to using Cosmos Transfer for photorealistic domain augmentation. The generated data will support training Vision-Language-Action (VLA) models and be delivered to frontier lab clients. This is a hands-on role involving simulation, trajectory generation, visual augmentation, dataset curation, and policy benchmarking. The intern will also document pipelines and create technical guides so the team can reproduce and scale the work. The internship is listed as lasting 3–6 months and is located in Mountain View, California.

Requirements

  • Currently pursuing or recently completed an MS or PhD in Robotics, Computer Science, Mechanical Engineering, or a related field.
  • Hands-on experience with at least one physics simulator: NVIDIA Isaac Sim/Isaac Lab, MuJoCo, PyBullet, or Gazebo.
  • Strong Python skills and comfort working in Linux/Docker environments with GPU workloads, including CUDA and multi-GPU scheduling.
  • Familiarity with robot learning fundamentals, such as imitation learning, behavior cloning, reinforcement learning, or sim-to-real transfer.
  • Working knowledge of USD/URDF scene descriptions and at least one robotics middleware, such as ROS/ROS2.

Nice To Haves

  • Prior experience with NVIDIA Omniverse, Isaac Sim extensions, or Replicator for synthetic data generation.
  • Familiarity with VLA architectures such as RT-2, Octo, π0, or GR00T N1, or open robot datasets such as Open X-Embodiment, DROID, or BridgeData.
  • Experience with domain randomization, photorealistic rendering, or neural rendering such as NeRF or Gaussian Splatting for sim-to-real.
  • Published research or course projects in robot manipulation, locomotion, or synthetic data for embodied AI.

Responsibilities

  • Build and operate synthetic trajectory pipelines using NVIDIA Isaac Sim, Isaac Lab, and the GR00T-Mimic blueprint to generate large-scale manipulation and locomotion datasets from a small number of human demonstrations.
  • Configure simulation environments, including scene composition, physics parameters, robot URDF/USD models, camera placements, and domain randomization, to improve trajectory diversity and sim-to-real transfer quality.
  • Apply NVIDIA Cosmos Transfer or GR00T-Dreams augmentation to transform simulation-rendered frames into photorealistic training images with varied lighting, textures, and backgrounds.
  • Design and run data quality experiments measuring success rates, trajectory smoothness, and visual fidelity, and iterate on pipeline parameters to improve downstream policy performance.
  • Curate and version datasets in formats compatible with VLA model training, such as Open X-Embodiment and LeRobot, ensuring metadata, task labels, and action annotations meet client specifications.
  • Benchmark synthetic versus real data by training imitation-learning policies, such as ACT and Diffusion Policy, on mixed datasets and reporting sim-to-real transfer metrics.
  • Document pipelines and author technical guides so the team can reproduce, scale, and extend the work.

Benefits

  • Founding-stage impact: pipelines will help define how Deccan AI generates physical AI data.
  • Exposure to frontier labs, with work consumed by clients such as Google DeepMind.
  • Full-stack ownership across simulation setup, trajectory generation, visual augmentation, dataset curation, and policy benchmarking.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service