XDOF, just 3 months out of stealth, is in talks for a Series B at a $1.2B valuation
The company, founded by UC Berkeley researchers Philipp Wu (CEO) and Fred Shentu (CTO), is moving toward a $1.2 billion valuation after a $70 million Series A in June that featured Andreessen Horowitz, Thrive Capital, Lux and Spark Capital. Revenue is now annualized at about $50 million, a growth rate that has prompted investors to approach XDOF despite its original plan to pause fundraising. XDOF’s core product is an end‑to‑end pipeline that uses low‑cost teleoperation hardware (the GELLO system) and human‑wearable sensors to capture real‑world manipulation data, which it packages for AI labs and robot builders that lack their own data‑collection infrastructure.
The move reflects a broader shift in the AI ecosystem: while large language models have long thrived on internet‑scale text corpora, general‑purpose robots still suffer from a “data desert.” XDOF positions itself as the robotics analogue of data‑labeling powerhouses such as Scale AI or Mercor, promising the “ABC” dataset—the largest curated collection of robot training footage to date. Competitors like Mecka AI are also courting the nascent market, and established labeling firms are extending into physical‑world data, underscoring a race to become the default data supplier for the next wave of embodied AI.
If the Series B closes, the infusion will likely fund rapid expansion of XDOF’s global collector network, accelerate the rollout of the ABC dataset, and deepen relationships with frontier AI labs that already count among the startup’s 20 customers. However, scaling a human‑intensive data pipeline brings operational risk: recruiting, training, and retaining teleoperators and sensor‑wearing annotators worldwide is costly and may encounter quality‑control challenges. Moreover, the valuation hinges on the assumption that demand for robot‑training data will keep rising as hardware catches up with software; any slowdown in robotics deployment could pressure the business model. Watch for the final raise amount, the terms of the 8VC lead, and whether XDOF can convert its early customer traction into recurring, high‑margin contracts.
Key Takeaways
XDOF’s projected $1.2 billion valuation comes less than three months after its Series A, driven by $50 million in annualized revenue.
The startup’s GELLO teleoperation system and sensor‑based collection workflow aim to create the first large‑scale, high‑quality robot training dataset, dubbed ABC.
Investors are betting that XDOF will become the “Scale AI for robotics,” a role that could define the data supply chain for embodied AI.
Scaling a human‑centric data pipeline presents execution risk; the next funding round’s size and the company’s ability to lock in long‑term customers will be key indicators of sustainability.
About the Source
This analysis is based on reporting by TechCrunch. Here is a short excerpt for context:
The round is being raised just months after the robot data startup exited from stealth.Read the original at TechCrunch
