Veeda AI is building the next generation of multimodal foundation world models for Physical AI. We're a small, fast-moving team of engineers and researchers from leading AI labs, tackling some of the most challenging problems at the intersection of AI, robotics, and embodied intelligence. If you're excited about pushing the boundaries of what's possible with Physical AI, you'll have the opportunity to make an outsized impact from day one.
Simulator integration: Help us connect our world models to existing simulator APIs that customers are familiar with (Isaac Lab, Carla, etc).
Harness engineering: Rethink "classical" to become a physical AI equivalent of an LLM coding harness.
Synthetic data generation: Own the moment when the world model and policy close the loop, collect success/failure cases of both algorithms.
Scene & Asset Pipelines: Build procedural scene generation and articulated asset authoring on OpenUSD, URDF, and MJCF, with automated convex decomposition, inertial-property assignment, and the QA that stops a broken asset reaching a training run.
Throughput & Determinism: Help us keep rollouts replayable under a fixed seed and configuration so a regression can be bisected.
Environment API & Fidelity: Own the interface policy training runs against (Gymnasium-style APIs, Isaac Lab managers) and the regression benchmarks that score both task success and divergence from real-hardware traces.
Bachelor's degree or equivalent hands-on experience in Computer Science, Computer Graphics, Robotics, or a related technical field.
Experience building or substantially extending a physics engine or simulation framework (Newton, MuJoCo, PhysX, or something proprietary).
Strong C++ and Python skills and experience writing GPU code in CUDA, NVIDIA Warp, or an equivalent data-parallel framework.
Fluency in 3D content pipelines (OpenUSD, glTF, meshes, materials, coordinate frames, units, inertias) and their intricacies.
Ability to measure simulation fidelity and close quantified gaps between simulators and the hardware they model.
Development experience with OpenUSD tooling or Omniverse Kit extensions.
Scalable procedural content generation skills (Infinigen, Houdini, Blender Geometry Nodes).
Experience with differentiable simulation or gradient-based system identification.
Experience building 3D Gaussian splatting or NeRF reconstruction pipelines with simulatable outputs.
Sensor modeling skills using real device characterization data (lidar, radar, event cameras, tactile arrays).
Contributions to open-source simulation projects (Newton, MuJoCo Warp, Isaac Lab, mjlab).
