Two rounds of financing secured in 40 days, the pragmatic player in the embodied data track enters the market.

The battle for embodied intelligence has extended to the data sector.

PE Daily learned that today (August 12), SCALEFORCE, an embodied intelligence data infrastructure company, announced its new round of financing for the first time: it completed two rounds of financing in 40 days, with investors including top domestic embodied intelligence industry players, Hengxu Capital and Kailian Capital.

Founded less than three months ago, SCALEFORCE has rapidly entered the operation stage: a multi-million-level data order from its lighthouse customer Shitou Intelligent Navigation has been fully launched, and it has successively established cooperation with multiple world model companies, leading embodied intelligence ontology manufacturers and industry solution providers... The commercialization flywheel is accelerating its operation.

"The competition of embodied intelligence has entered the second half stage focused on data and intelligence." This is the real situation Guo Jiangliang saw, and it is also the reason why he chose to start his own business. As he said, what SCALEFORCE aims to do is to enable data to run through the whole lifecycle of physical AI, and generate infinite intelligence in the physical world.

The hotter embodied intelligence gets, the scarcer data becomes

The grand occasion is still vivid in our minds — looking back at the first half of 2026, the total financing amount of China's embodied intelligence track reached 935 billion yuan, surging 5 times over the same period of last year.

However, the hotter embodied intelligence gets, the scarcer data becomes. A widely recognized consensus in the industry is that the embodied large model capable of general autonomous capabilities requires at least 10 million hours of high-quality real interactive data. By the beginning of 2026, the total amount of available high-quality physical interaction data worldwide was only about 500,000 hours — the gap exceeds 99%.

Different from large language models that can crawl training materials from the Internet, embodied intelligence requires multi-modal physical interaction data with aligned vision, touch, joint trajectory, object mechanics and environmental timing — these data cannot be obtained online and must be collected in the real world.

What is more tricky is that the industry is still facing three dilemmas: uncontrollable data quality — low spatial-temporal alignment accuracy of multi-modal data, and abnormal conditions cannot be monitored in real time; insufficient data scale — collection equipment cannot be managed on a large scale, with low automation level; poor data generalization ability — severe OOD problem, and cross-ontology reuse is almost impossible.

This is exactly the opportunity Guo Jiangliang saw when he founded SCALEFORCE — the competition of embodied intelligence is essentially the competition of data. Whoever can build barriers in data infrastructure will take the initiative in this competition.

Through SCALEFORCE, we see an answer: redefining the infrastructure of physical AI from the data source.

At present, SCALEFORCE has developed the MatrixOS, a physical AI operating system. Among them, the ADA (Action-Data Alignment) data generalization engine is designed to solve the problem of cross-ontology data reuse. According to the real machine multi-scenario evaluation, its success rate has jumped from 65% to 92%.

On the other side is data production efficiency. SCALEFORCE has independently developed the GDP (Global-Deep-Proactive) data quality engine, which continuously optimizes data collection strategies through information entropy analysis, posterior probability estimation, action semantic analysis and Bayesian active learning, achieving an available data conversion rate of 80%, the highest level in China.

At the data collection end, SCALEFORCE has built a global large-scale physical AI data production network, which supports unified management and scheduling of hundreds of thousands of level data collection nodes worldwide, with the full-process automation degree exceeding 95%. In addition, SCALEFORCE adheres to the human-centric data collection paradigm, and has taken the lead in developing a data collection kit with sub-millisecond multi-sensor time synchronization worldwide, which can synchronously collect full-modal data such as vision, touch, force sense and behavior on a large scale.

In simple terms, MatrixOS is more like a "production system" built around embodied intelligence data: the front end connects to the real physical world to continuously obtain data; the middle part completes data processing, cleaning, generalization and management; the back end delivers data to models, robots and specific application scenarios.

Once this link is fully operational, data will no longer be a one-time project delivery, but can continuously enter the next round of training to generate new data, and finally form a data flywheel. It is reported that SCALEFORCE is building a high-quality data production and distribution network with the largest global production capacity and the lowest cost. Once the network is completed, the scale effect and cost advantage of data will become the strongest competitive barrier.

Industry Window Period

Just Won a Multi-Million Level Order

When will robots really enter thousands of households?

At the current stage, the popularity of embodied intelligence is beyond doubt. As the industry begins to enter the large-scale development stage, data collection, cleaning, generalization, management and distribution are gradually becoming an independent industrial link.

This is exactly the position SCALEFORCE has chosen. Based on its fully developed MatrixOS platform, SCALEFORCE can provide targeted scenario data for ontology manufacturers, real physical interaction data for world model companies, and build data closed loops for industry customers around specific tasks.

This neutral positioning allows SCALEFORCE to undertake the strong demand from all links of the industrial chain. At present, the company has won a multi-million-yuan data order from Shitou Intelligent Navigation — through the service of "targeted scenario data collection + MatrixOS platform data processing", it helps Shitou Intelligent Navigation's A-series robots successfully settle in multiple operation scenarios of Aptiv's factories.

At the same time, SCALEFORCE has established in-depth cooperative relations with multiple world model companies including Zhizai Wujie, leading embodied intelligence ontology manufacturers and industry solution providers; it has also joined the Huawei Ascend computing ecosystem, and is one of the first batch of global open source embodied intelligence/world model data and algorithm pipeline contributors of Ascend. In addition, the company has established in-depth cooperation on cutting-edge exploration of embodied intelligence with Peking University and Beihang University, to jointly explore the technical boundary of next-generation embodied intelligence.

In response to the demand for high-quality real scenario data from overseas markets, SCALEFORCE's international layout has also been substantially launched, and it has formed an overseas business team led by senior Southeast Asian operation and sales experts to promote the construction of overseas market business closed loops and partner networks.

However, the outside world is still curious: how can a company founded less than three months ago achieve such a fast development speed?

Although it is a new player, SCALEFORCE is backed by a team with profound accumulation — its founder Guo Jiangliang is a founding member of Baidu Intelligent Cloud, who incubated multiple core products from scratch including Baidu Cloud MapReduce, machine learning platform, enterprise AI middle platform and industrial quality inspection cloud, led the team to achieve business breakthroughs in multiple strategic industries such as industry, finance, energy and power, and accumulated hundreds of millions of yuan in revenue.

Later, he served as Vice President of Technology of AInnovation, an AI+ manufacturing enterprise, and fully built the technical system of industrial large model and industrial embodied intelligence, experiencing the whole process of the company from startup to listing. Throughout his career, Guo Jiangliang's most distinctive label is that he has been deeply engaged in the forefront of Data&AI Infra technology and business.

Chief Scientist Alex, Ph.D. from the Institute of Computational Linguistics of Peking University, once served as a core member at Meta AI and Huawei Noah's Ark Lab successively, leading the design of China's first generation of trillion-parameter-level MoE architecture LLM, participating in the whole process from data formula, code development to underlying operator optimization, with profound cutting-edge algorithm technology reserve and international vision.

In addition, Victor, Chief Revenue Officer, once led the enterprise to achieve sales performance of nearly 1 billion yuan, with in-depth accumulation in China's AI, large model and embodied intelligence ecosystem; Angel, Chief Product Architect, has led the design, R&D and implementation of multiple sets of enterprise-level data platform products.

With the three pieces of puzzle of technology, product and commercialization assembled, a practical team that has known and worked together for many years, combining cutting-edge technical vision and commercial operation experience has taken shape, which becomes the foundation of SCALEFORCE's product capability. In Guo Jiangliang's words, "The biggest feature of our team is that we have fought side by side in the past, and we are all industrial practitioners and serial entrepreneurs who are experienced in data operation and performance growth."

At present, the timing is becoming more and more critical. As this year is widely recognized by the industry as the "first year of embodied data", Guo Jiangliang led the team to enter the market decisively. The industry generally believes that 2027 to 2029 may become an important window period for the large-scale commercialization of humanoid robots. If robots are deployed on a large scale in the next few years, the data infrastructure that seems to be in the early stage today will also usher in a real explosion.

The wave of development is coming. Unitree Technology's IPO on the Sci-Tech Innovation Board is around the corner, and there is a long queue of companies waiting for IPO behind it. After several years of development, embodied intelligence has finally reached the moment of value realization. The next stop is the time window for embodied data that cannot be missed.

This article is from the WeChat Official Account "PE Daily" (ID: pedaily2012), written by Wu Qiong, authorized for release by 36Kr.