Senior Asset Pipeline Engineer

Responsibilities

• Design the OpenUSD schema and composition strategy — references, variants, payloads, custom schemas — for a large, versioned 3D asset library that must scale across authoring tools and simulation engines
• Build and own automated pipelines that move assets between DCC tools and real-time/simulation engines, removing manual hand-offs
• Define automated validation gates that catch asset-integrity and fidelity issues in CI, before they reach production
• Solve material conversion for simulation: translate complex authored materials into USD-native representations (UsdShade, and MDL where needed) while preserving the physical fidelity sensor simulation requires
• Design the asset metadata strategy that powers discovery, semantic labeling, and downstream tooling
• Take the pipeline from design through production — observability, reliability, and a smooth experience for the content authors who depend on it

Requirements

• Deep, practical OpenUSD composition expertise — references, variants, payloads, inherits, specializes — designing and debugging composition stacks
• Experience authoring and validating custom USD schemas (API and applied schemas)
• Strong UsdShade material expertise: authoring, validating, and converting complex material networks
• Built and owned automated DCC-to-engine pipelines in production — version-controlled Python that runs in CI, with real downstream users
• Managed USD at library scale: versioning, identifier and namespace conventions, and validation across thousands of assets

Nice-to-Haves

• NVIDIA Omniverse / Isaac Sim: Kit extensions (Python or C++), USD ingest, and Replicator / synthetic-data-generation pipelines
• Houdini LOPs: native USD authoring and procedural scene graphs
• MDL (Material Definition Language): physically-based material authoring and portability across renderers
• Sensor simulation domain knowledge: how material properties (reflectance, roughness, emissivity) drive lidar, radar, and camera outputs
• Automotive, defense, or robotics simulation experience: contexts where sensor fidelity matters beyond visual aesthetics (ROS 2 / scene assembly a plus)