WeRide subsidiary JingShuo Technology completes tens of millions of USD pre-A round financing

On October 29, WeRide (NASDAQ: WRD, HKEX: 0800) disclosed the first financing progress of its subsidiary JingShuo Technology. JingShuo Technology has now received a pre-A round of joint investment totaling tens of millions of US dollars from investors including Jiayu Capital, HKSTP Hong Kong Science and Technology Parks, and Sirius Capital.

JingShuo Technology is committed to building the data and model infrastructure for the physical AI era. With the core philosophy of "high-quality data driving the evolution of physical AI," and relying on the integrated data-model platform WorldEngine and the self-developed world model GENESIS-Robotics, it has formed a product system covering the self-developed first-person-view high-precision data collection device EGOK, large-scale data collection, high-quality datasets, and customized data services, connecting real-world operational experience with the robot learning process.

At present, JingShuo Technology has established cooperation with dozens of the world's top large-model companies and embodied intelligence enterprises, with 100+ customers and ecosystem partners, and orders in hand amounting to several billion yuan. It provides strong data support for the R&D and application of physical AI and general large models, with annual revenue growth exceeding 100%, demonstrating strong commercialization capabilities and growth potential.

As artificial intelligence accelerates from the digital world into the physical world, high-quality data and related infrastructure are becoming key elements driving technological evolution and industrial development. It is widely believed in the industry that whether for large language models, world models, or embodied intelligence systems, capability improvements are built on the continuous accumulation, governance, and iteration of high-quality data. Especially in the field of physical AI, the collection, processing, validation, and training of real-world data have become an important foundation determining the boundaries of model capabilities.

As one of WeRide's businesses, JingShuo Technology relies on WeRide's nearly ten years of experience in commercializing physical AI to continuously strengthen its capabilities in world models, data platforms, and skill asset engines, building a complete infrastructure system covering data production, data governance, model training, and capability accumulation, providing underlying support for next-generation physical AI systems.

The core team of JingShuo Technology comes from WeRide's physical AI R&D system. CEO Huo Da is WeRide's Employee No. 001 and has long been deeply engaged in autonomous driving platforms and physical AI data infrastructure construction. He has led the establishment of global engineering R&D teams in the United States, China, and Singapore, and has extensive experience in technical architecture and data system construction.

After this round of financing, JingShuo Technology will further expand its core R&D team, deepen the synergistic capabilities of its data platform, world model, and skill asset engine, enrich the task and scenario coverage of high-quality datasets, and accelerate the construction of infrastructure capabilities for the physical AI field.

From a global perspective, the AI industry is currently developing rapidly, and data infrastructure companies continue to be favored by the capital market. In June 2025, US tech giant Meta made a strategic investment of US$14.3 billion in AI data platform Scale AI, corresponding to a company valuation of US$29 billion; in October of the same year, Mercor announced the completion of a US$350 million Series C financing at a valuation of US$10 billion; in September this year, Snorkel AI completed a US$350 million Series E financing at a valuation of US$3.5 billion. Domestic physical AI data company Guanglun Intelligence has reached a valuation of US$2 billion and completed a new round of 1 billion yuan in strategic financing in June this year.

This article is provided by JingShuo Technology. QbitAI is authorized to republish it, and the views belong to the original author.