AI robotics platform with simulation, vision, and manipulation tools
Intrinsic is Google's robotics group building tools and frameworks for industrial automation. The stack reveals a simulation-first architecture (Gazebo, MuJoCo, Isaac Sim) paired with modern ML infrastructure (JAX, PyTorch, TensorFlow on GCP TPUs/GPUs), suggesting heavy investment in physics-based training and sim-to-real transfer. Engineering dominance in hiring (39 of 47 open roles) and projects centered on grasping algorithms, real-time frameworks, and robotics data pipelines indicate a company in active platform expansion rather than sales-led scaling.
Notable leadership hires: Software Engineering Lead
Intrinsic develops an AI robotics platform designed to make industrial automation accessible to a broader set of businesses and developers. The company operates across three layers: simulation environments (Gazebo, MuJoCo, Isaac Sim), machine learning for perception and control (3D vision models, grasping optimization), and deployment infrastructure (Kubernetes, real-time frameworks, distributed data pipelines). The organization is structured around algorithm development and infrastructure engineering, with active projects spanning robot manipulation, computer vision, data pipelines for model training, and tools for platform accessibility. Headquarters in Mountain View with hiring across the US, Switzerland, Germany, and Singapore.
Python, C++, JAX, PyTorch, TensorFlow, Gazebo, MuJoCo, Isaac Sim, ROS/ROS 2, Kubernetes, GCP infrastructure, gRPC, and Three.js/WebGL for visualization. Heavy simulation and ML compute on TPUs and GPUs.
Headquartered in Mountain View, California. Actively hiring in the United States, Switzerland, Germany, and Singapore.
Other companies in the same industry, closest in size