XDOF, a startup building real-world teleoperation data pipelines to train general-purpose robots, is in late-stage talks to raise a Series B at a valuation of roughly EUR 1 billion, according to people familiar with the deal. The round is being led by 8VC and comes less than three months after the company exited stealth.
The pace of the deal stands out. XDOF closed a EUR 60 million Series A in June, with backing from Thrive Capital, Andreessen Horowitz, Lux, and Spark Capital, and had not planned to raise again so quickly. Rapid expansion changed those plans, with annualized revenue approaching EUR 43 million prompting venture investors to approach the company about a new round.
The total amount being raised and whether the valuation includes the new capital could not be confirmed. Deal terms are not final and may still change. XDOF and 8VC did not respond to requests for comment.
What XDOF Builds
XDOF was co-founded in 2024 by UC Berkeley researchers Philipp Wu, who serves as CEO, and Fred Shentu, the CTO. The company aims to build the data pipelines, collection tools, and annotation systems that frontier AI labs and robotics companies struggle to create on their own, effectively serving as an outsourced data-supply chain for the robotics industry.
The concept grew out of Wu’s PhD research on how robots learn from large datasets, where a shortage of large-scale data proved a major obstacle. Together with Shentu, he developed GELLO, a low-cost teleoperation system that lets a human operator control a robotic arm remotely to generate training data. Their work produced an influential robotics paper that became the foundation for XDOF.
Solving the Robotics Data Bottleneck
Investors describe XDOF as the Scale AI or Mercor for physical robotics, a nod to the data-labeling firms that helped power the AI boom. Unlike large language models, which trained on vast amounts of internet content, physical robots lack an equivalent real-world dataset, making data collection a critical bottleneck for building general-purpose machines.
The company is partnering with UC Berkeley’s AI Research lab to release what it says is the largest collection of high-quality robot training data ever assembled, called ABC. To gather this data, XDOF combines remote robot teleoperation with human collectors who wear sensors to record everyday tasks such as folding clothes and flattening boxes.
XDOF plans to hire and train teams of data collectors worldwide, including teleoperators who steer robots remotely and egocentric operators who wear body sensors to capture movement. The startup has said it is already working with 20 customers, several of them frontier AI labs. Other companies pursuing real-world robot training data include Mecka AI, along with human-data platforms expanding beyond language models such as Scale AI and Micro1.
Source
Image: techcrunch.com
