Just three months after emerging from stealth, XDOF is already in late-stage talks to raise a Series B at a staggering $1.2 billion valuation. This remarkable trajectory underscores the immense demand for robust robotic data collection and infrastructure in the AI industry.
The potential round, reportedly led by 8VC, comes on the heels of a $70 million Series A in June. That earlier round saw participation from heavyweights like Thrive Capital, Andreessen Horowitz, Lux, and Spark Capital. While the exact amount of new capital and whether the valuation is pre- or post-money remain unclear, the speed of this follow-on is telling. The valuation is currently in flux and could still change.
Why VCs Are Eager to Invest
The startup wasn’t planning to raise again so soon. However, XDOF’s explosive growth—with annualized revenue approaching $50 million—prompted investors to approach the company. This rapid financial momentum highlights a critical bottleneck in the robotics sector: the need for high-quality training data. The company’s ability to generate revenue at this scale, less than three months out of stealth, is exceptional even by Silicon Valley standards.
XDOF and 8VC didn’t respond to requests for comment, but industry observers note that such rapid fundraising typically signals strong customer demand and market validation. The startup’s trajectory suggests that the robotics industry’s hunger for reliable training data is far outpacing current supply.
Solving the Robotics Data Bottleneck
XDOF focuses on a core problem. Unlike large language models that can train on vast swaths of the internet, physical robots lack an equivalent digital repository. This makes robotic data collection a fundamental obstacle to building general-purpose machines that can operate effectively in the real world.
Co-founded in 2024 by UC Berkeley researchers Philipp Wu (CEO) and Fred Shentu (CTO), XDOF aims to build the data pipelines, collection tools, and annotation systems that frontier AI labs and robotics companies can’t easily build themselves. Essentially, they act as an outsourced data-supply chain for the entire robotics industry. Without companies like XDOF, individual robotics firms would need to invest heavily in their own data infrastructure, which is both costly and time-consuming.
From Research to Real-World Application
The company’s foundation lies in academic research. While a PhD student, Wu faced a significant impediment: the lack of “large-scale data to work with.” He teamed up with Shentu on a project called GELLO. This low-cost teleoperation system allows human operators to control robotic arms remotely to generate training data. Their work produced an influential paper in robotics, which now serves as the foundation for XDOF.
This academic pedigree gives XDOF credibility in a field where technical expertise is paramount. The transition from research to commercial application has been remarkably swift, demonstrating the founders’ ability to execute on their vision while maintaining scientific rigor.
How XDOF Collects Data
XDOF is now actively building on that research. To capture the necessary data, the company uses a combination of methods.
Remote Teleoperation: Operators steer robots remotely to perform specific tasks, generating precise movement data.
Egocentric Data Capture: Human collectors wear sensors to record everyday activities, capturing natural human movements.
This dual approach allows them to gather a wide range of data, from folding clothes to flattening boxes. The data is then used to train more capable and adaptable robots. The combination of teleoperation and egocentric capture provides a comprehensive dataset that reflects both controlled and naturalistic movements.
The “Scale AI for Robotics” Model
Investors often describe XDOF as the Scale AI or Mercor for physical robotics. This comparison references the data-labeling giants that fueled the AI boom by providing the structured data needed to train advanced models. XDOF aims to play the same role for the robotics industry.
The company is already making significant strides. XDOF is partnering with UC Berkeley’s AI Research lab to release what it believes is the largest collection of high-quality robot training data ever assembled, dubbed ABC. This partnership leverages the university’s research expertise while giving XDOF access to top-tier talent and facilities.
Global Expansion and Customer Base
To meet the growing demand, XDOF plans to hire and train teams of data collectors worldwide. This includes teleoperators who steer robots remotely and egocentric operators who wear body sensors to capture movement data. The global nature of this operation allows XDOF to collect diverse data from different environments and cultures, creating more robust and adaptable training datasets.
XDOF previously told TechCrunch that it is already working with 20 customers, including several frontier AI labs. This solid customer base adds weight to the company’s valuation claims and demonstrates immediate market fit. Working with frontier AI labs is particularly significant, as these organizations are at the forefront of developing advanced AI systems and have the most demanding data requirements.
Competition in the Data Collection Space
XDOF is not alone in this pursuit. Other startups attempting to collect real-world data for robot training include Mecka AI, as well as human-data platforms expanding beyond LLMs, such as Scale AI and Micro1. However, XDOF’s rapid growth and focus on physical robotics give it a distinct first-mover advantage.
The competitive landscape is evolving quickly, but XDOF’s combination of academic credibility, strong revenue growth, and blue-chip investor backing positions it well for long-term success. The company’s ability to raise funding at such a high valuation so soon after its Series A suggests that investors believe XDOF has a sustainable competitive advantage.
The Future of Robotics Depends on Data
As the robotics industry continues to grow, the demand for high-quality training data will only increase. XDOF’s success in robotic data collection positions it as a critical infrastructure provider for the AI-driven future. The company’s potential $1.2 billion valuation reflects investor confidence that data collection will remain a bottleneck for years to come.
The partnership with UC Berkeley, the planned global expansion, and the growing customer base all point to a company that is scaling effectively. If the Series B closes as expected, XDOF will have the capital needed to accelerate its data collection efforts and solidify its market position.
XDOF’s rapid ascent from stealth to a potential $1.2B unicorn status highlights the critical importance of data infrastructure in the AI-driven future. By solving the complex challenge of robotic data collection, XDOF is positioning itself as an indispensable partner for the next generation of robotics companies and frontier AI labs. As the deal potentially closes, it will be a significant indicator of where the smart money sees value in the AI ecosystem.
The startup’s journey from academic research to commercial success in just two years is a testament to the founders’ vision and execution. For those watching the robotics and AI space, XDOF represents one of the most compelling stories of 2026—a company that identified a critical gap and moved quickly to fill it.
