Rapid Rise to Unicorn Status
XDOF, a startup specializing in robotic training data, is currently in late-stage talks for a Series B financing round. If finalized, the deal will value the company at $1.2 billion. This development arrives less than three months after the firm exited stealth mode. Sources familiar with the negotiations indicate that 8VC is leading the round. The speed of this valuation growth surprised observers, especially given the company closed a $70 million Series A round in June 2026.
Investors appear motivated by the company's aggressive revenue trajectory. Current annualized revenue figures sit near $50 million. While terms of the pending Series B remain in flux and the exact amount of capital is undisclosed, the potential valuation signals a massive investor appetite for robotics infrastructure. XDOF representatives and 8VC have not provided comments regarding these reports.
Solving the Physical Data Bottleneck
Founders Philipp Wu and Fred Shentu established XDOF in 2024. Both founders arrived at the startup from UC Berkeley. Their primary objective involves building the pipelines and annotation tools necessary for general-purpose robot development. The robotics industry suffers from a shortage of real-world training data. Unlike software models that rely on internet-scale text datasets, robots require high-fidelity movement data. XDOF occupies this gap by acting as a specialized supply chain for data collection.
Wu previously encountered this research hurdle as a PhD student. The project GELLO, a low-cost teleoperation system, served as the genesis for the company. By allowing human operators to control robotic arms remotely, the team generated the data foundational to their business. This system allows for the capture of mundane tasks like folding clothes or flattening boxes. Investors now compare the firm to data-labeling titans like Scale AI, viewing it as a prerequisite for the next wave of mechanical automation.
Global Operations and Competitive Positioning
XDOF currently maintains a partnership with UC Berkeley’s AI Research lab to distribute the ABC dataset, which they claim is the largest collection of its kind. The company employs human collectors across the globe. These operators use sensors to record human movement and steer robots from remote locations. With 20 customers already on the roster, including several prominent AI laboratories, the firm is scaling its physical footprint rapidly.
Competition in this space is heating up. Other players such as Mecka AI and established platforms like Micro1 and Scale AI are also chasing the market for robotics data. XDOF holds a narrow head start by focusing specifically on the intricacies of physical teleoperation. The coming months will determine if the company can maintain this pace as it expands its global network of human collectors. Market analysts will watch for further details on the Series B closing, as it will likely define the funding environment for robotics infrastructure for the remainder of the year.