Spatial AI for generating and interacting with 3D worlds
World Labs builds Large World Models (LWMs) that perceive, generate, reason about, and interact with 3D environments from multimodal inputs like text, images, and video. They serve industries including gaming, robotics, film, architecture, and design by enabling persistent, editable 3D worlds. This advances AI beyond 2D to spatial intelligence, unlocking new applications in simulation, creativity, and physical world understanding.
Owns developer infrastructure including Bazel-based monorepo build systems, test infrastructure, and release pipelines to enhance engineering velocity and reliability. Requires 5-10 years experience with build tools like Bazel and languages like Python, Go, or Rust.
Builds mission-critical training, data, and inference infrastructure for AI research on large world models. Optimizes performance, debugs distributed systems, and improves researcher velocity with 5+ years experience in ML systems and strong Python proficiency.
Develops and trains large-scale diffusion models for generating 3D worlds, incorporating novel control signals and integrating cutting-edge generative AI research into production. Requires 3+ years in generative modeling, PyTorch/TensorFlow expertise, and strong publication record.
Owns design, architecture, and operation of database and storage systems for high-throughput, low-latency AI platform workloads. Requires 8+ years experience with transactional databases like Spanner, Postgres; expertise in data modeling, distributed systems, and production reliability.
Senior Product Engineer owns end-to-end product features as a full-stack developer, partnering with product leadership on roadmapping and execution while mentoring teams. Requires 4-10 years experience with TypeScript, React, and web architecture, plus product-minded approach.
Designs, implements, and advances state-of-the-art SLAM systems for robust spatial understanding from sensor data in dynamic environments. Requires 6+ years in SLAM/state estimation, proficiency in Python/C++, and expertise in probabilistic methods and sensor fusion.
Leads technical vision and execution for 3D modeling and reconstruction systems, driving innovation in learning-based methods, mentoring a small research team, and delivering production-scale capabilities integrated with rendering platforms. Requires 8+ years in computer vision/graphics with strong Python/C++ and ML expertise.
Develops state-of-the-art 3D reconstruction systems from real-world data using feed-forward and optimization-based methods. Requires 6+ years in computer vision/graphics/ML, proficiency in Python/C++/PyTorch, and expertise in 3D representations.
Research Engineer improves world models by curating multimodal (image/video) training datasets at scale. Builds processing pipelines, deploys ML for enrichment, and closes data-model-evaluation loop using strong engineering and research skills.
Builds production pipelines integrating 3D graphics/ML research into reliable backend services and APIs for a web app focused on Gaussian Splat scene generation/editing. Requires strong Python, backend skills, 3D ecosystem fluency, and experience productionizing prototypes.