Jie Junyuan, former head of embodied intelligence at Qianxun, founded embodied data collection company gi labs in North America

Jie Junyuan

It is reported that Jie Junyuan, former head of the embodied intelligence department at Qianxun, has started a business in North America, founding gi labs (General Intelligence Lab). The company mainly focuses on building data collection devices, including motion capture gloves, multi-camera setups, robotic grippers, and more. Among them, the EGO1GS integrated data collection headset is equipped with 1080p binocular cameras with a field of view up to 157H 84D, and a sampling frequency twice that of the previous generation, at approximately 400Hz.

According to the aforementioned informed sources, gi labs currently has an overseas office in San Francisco, USA, with its main domestic office in Suzhou. The team consists of AI researchers, roboticists, and engineers. The company is currently hiring mechanical engineers and developing products such as robotic grippers, multi-camera setups, and motion capture gloves.

At the product level, gi labs is currently refining a new generation of data collection devices. Informed sources say the product is still in the prototype stage and is expected to enter mass production in early November; pricing starts at 3,555 yuan per set, with 3,150 yuan per set for orders of 1,000 or more. Compared with the previous generation, the core differences of the new product are a frame rate increased to 60fps and an improved field of view (FOV).

Public information shows that Jie Junyuan graduated with a bachelor's degree from the University of Science and Technology of China. In his junior year, he published a paper on image denoising and restoration at NeurIPS as first author, with over 1,900 citations. He then went to the University of Washington in the United States to pursue a doctorate. His career spans Amazon, ByteDance, and Qianxun Intelligence. He led the development of the MXNet system, was responsible for ByteDance's recommendation system and federated learning platform, and from 2025 to 2026 was fully responsible for advancing Qianxun Intelligence's Spirit VLA large model, promoting its deployment in scenarios such as CATL and JD.com. In February 2026, Qianxun Intelligence completed two rounds of financing totaling nearly 2 billion yuan, with a valuation exceeding 10 billion, and investors including HongShan China, Yunfeng Fund, Chaos Investment, and industrial capital such as CATL. Jie Junyuan previously held a 15% stake in its controlled subsidiary Wanjing Qianxun and exited in March of this year. In April 2026, he formally left Qianxun.

Currently, there is also a large supply-demand gap for embodied data. According to estimates by the China Academy of Information and Communications Technology, for embodied models to achieve a breakthrough like ChatGPT, at least tens of millions of hours of effective data support are needed, while currently only millions of hours of high-quality data are available globally.

However, for Jie Junyuan and gi labs, credentials and capital are only entry tickets. Mass production and delivery in November and repeat purchases by the first batch of customers will provide the first hard answer.