XDOF Raises Series B at $1B Valuation Months After Stealth Exit
In brief
- XDOF closed $70M Series A and is raising Series B at $1B valuation
- Data pipelines and teleoperation tools for robotic systems training
- Founded October 2024, emerged from stealth June 2026 with 60 employees
- Released ABC-130K open-source bimanual robot dataset with UC Berkeley and MIT
Rapid Scaling in Data Infrastructure
XDOF's swift move toward Series B reflects intense market demand for high-quality training data pipelines in robotics and AI. The company currently employs around 60 people and counts approximately 20 customers, mostly frontier AI labs and robotics firms. Total funding to date sits around $78 million when you include an earlier seed round.
Founded in October 2024 by Philipp Wu (CEO), Yide (Fred) Shentu, and Nemo Jin, XDOF moved from concept to meaningful traction in less than two years. The founders built a platform addressing a critical bottleneck: collecting and labeling the massive datasets required to train modern robotic systems.
What XDOF Builds
The company builds data pipelines, teleoperation tools, and full-stack tooling for collecting and processing the high-quality training data that advanced robotic systems need to learn how to do things in the physical world. This infrastructure sits between raw sensor data and model training, a layer many robotics startups struggle to build in-house.
XDOF's approach has attracted both venture capital and academic partnerships. The company has partnered with UC Berkeley and MIT to create and release ABC-130K, an open-source dataset for bimanual robot manipulations. The dataset contains 130,000 trajectories across 195 bimanual tasks, along with additional simulation resources.
Market Timing
The rapid funding cycle reflects investor confidence in XDOF's position within a growing robotics ecosystem. As embodied AI systems mature, the demand for specialized data infrastructure is outpacing supply. XDOF's early customer base and technical team suggest the company has found a defensible niche in a crowded landscape.
