GenRobot AI
Genrobot provides physical AI data infrastructure for embodied intelligence companies, offering high-volume, multi-modal data collection and processing services. The company's proprietary DAS hardware line captures human-centric interactions—including egocentric video, hand movements, and tactile signals—across household, manufacturing, and logistics scenarios, which are then processed through its Data Foundation Model into training-ready datasets for robot learning.
- Artificial Intelligence
- Robotics
- Hardware
Funding
Raised to date
$200MRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
across 4 rounds
AGBBBCI+10$100M · Series A Extension
Announced Nov 2026
Disclosed cumulative: $200M
Founders
Founder details are not available yet.
Product
Problem
Training general-purpose embodied AI systems requires massive amounts of high-quality, multi-modal data that captures real human interactions with the physical world. Existing datasets are often limited in scale, diversity, or annotation quality, slowing model iteration and hindering the development of robots capable of operating in unstructured real-world environments.
Solution
Genrobot provides a full-stack physical AI data infrastructure that turns large-scale human experience into training-ready datasets for embodied intelligence. The company's proprietary DAS hardware line—including Ego, Dex, Fingers, Gripper, and Controller—captures synchronized multi-modal data such as egocentric video, hand and full-body motion, tactile signals, and audio across diverse real-world scenarios. This raw data is processed through the Gen Matrix Governance Platform and a Data Foundation Model (DFM) that generates structured training signals including 3D hand reconstruction, depth information, and hierarchical semantic annotations. The company offers tiered data services ranging from raw collection to fully model-processed datasets, enabling robotics companies to accelerate their model development cycles.
Target Audience
Primary customers are embodied intelligence companies, robotics manufacturers, and AI research teams developing general-purpose robots for household, manufacturing, and logistics applications.
Features
- DAS Ego headset with six high-resolution RGB cameras, 270° horizontal FOV, and millimeter-level pose trajectory reconstruction via self-developed BEV-odometry neural network
- DAS Fingers and Dex capture tactile and manual precision data with magnetic encoders and agile touch sensing
- Data Foundation Model (DFM) converts multi-view video into synchronized 3D hand meshes, MANO parameters, large-FOV depth, and hierarchical task annotations
- Open datasets including Gen-HumanEgo with 1,800+ hours of first-person demonstrations across 10,000+ tasks
- Tiered data services: Raw Data (Pro), Processed Data (Max), and Model-processed Data (Ultra) with atomic annotations and 3D/2D segmentation
- Comprehensive skill taxonomy spanning 8 domains, 15 scenarios, 38 scenes, and 165 tasks across home, manufacturing, logistics, and service environments