5Y News | Data Company "Intelligent Knowledge" Completes Angel Round

Recently, Human Intelligence completed its angel round, co-led by Jinqiu Fund and Yaotu Capital. The proceeds will go toward two priorities:

Expanding frontier data categories: Deepening work in high-value data types such as coding, Enterprise Office (GDPVal), and agentic tool use, while actively exploring new scenarios including AI4Math, AI4Science, and AutoResearch;

Upgrading its expert network: Continuously refining expert profiling, skill assessment, and talent-role matching mechanisms to build a high-quality, highly engaged expert collaboration community.

5Y Capital commented: "The essence of machine learning is 1. obtaining better, more comprehensive data, and 2. modeling that data more efficiently. Looking back at the 'Bitter Lesson,' architectural improvements have always been incremental, while data-driven innovation can produce outsized impact. Building evaluation and RL environments represents the highest-leverage, most durable use of human time."

As large models continue to iterate, high-quality expert data is becoming the critical variable that determines model ceilings.

The United States has already produced leading expert data platforms such as Surge AI, Mercor, and Scale AI. By comparison, China still lags significantly in systematic capabilities in this domain, particularly in efficiently organizing cross-industry domain experts.

China has the world's largest pool of high-caliber talent — engineers, researchers, white-collar professionals, and experts across industries, vast in number and exceptional in quality. Chinese large models should not remain dependent on distilled data and overseas sources in the long run; they must establish a structural data advantage rooted in domestic expert intelligence. How to crystallize this professional expertise into high-quality data is a foundational imperative for Chinese large models competing globally.

Human Intelligence was founded precisely against this backdrop. The company's mission is unambiguous: Frontier Data, Made in China — helping Chinese models advance continuously and compete at the SOTA level.

In Human Intelligence's view, data quality determines AI's ceiling, and by extension, the heights this generation can reach. Rather than chasing benchmark scores alone, the team has written itself a different mission: "to cultivate AGI with genuine soul: curiosity, humor, creativity, and flashes of brilliance even humans didn't anticipate." In their eyes, data is not merely "training material" but "a profound act of nurturing" — coarse inputs cannot produce depth, and sheer volume cannot cultivate soul.

Human Intelligence operates two core businesses:

Human Intelligence (human-intelligence.cn): Focused on frontier large-model data production and methodology development;

Meetchances (meetchances.com): A leading large-model expert recruitment and collaboration platform in China, connecting experts across industries to participate in model training.

At the seed stage, Human Intelligence received investment from 5Y Capital and MiraclePlus. To date, Meetchances has facilitated hundreds of AI interviews daily and thousands of successful expert engagements monthly, with tens of thousands of experts registered on the platform.

The large-model competition unfolds not only at the algorithm and compute levels. High-quality data is equally an invisible infrastructure requiring sustained investment — this is the role Human Intelligence aims to play, one that cannot be ignored, in China's large-model industry.

"We firmly believe that transparent data production processes, democratized production tools, and scientific incentive mechanisms will transform the intellectual output of more Chinese experts into genuine global competitiveness for Chinese large models," the Human Intelligence team stated.

Finally, Human Intelligence continues to recruit. If you're curious about the high-quality data behind large models, or excel at managing complex expert collaboration projects at the scale of hundreds or thousands, they're eager to talk.