Former CICC and Oracle Team Launches Another Startup: Its AI Intelligent Automation Platform Penetrates 30% of Central State-owned Enterprises and Secures Hundreds of Millions in New Funding | Exclusive Report from Hardcore Kr
Author | Lin Qingqing
Editor | Yuan Silai
Yingke has learned that recently, "Jiuke Information", a provider of AI intelligent automation platforms, completed its Series B2 financing round. This round of financing was exclusively led by Shenzhen Special Zone Construction and Development Strategic Emerging Industry Private Equity Venture Capital Fund (hereinafter referred to as "Tejianfa Asia Merchant Fund"), with the cumulative financing amount exceeding 100 million RMB. Previously, the company had received multiple rounds of investment from investors including China Merchants Group Venture Capital, GSR Ventures, and Shenzhen Capital Group Sossou Fund. This round of financing will be mainly used for product R & D, market expansion, and overseas business layout.
Founded in 2020 and incubated within China Merchants Group, "Jiuke Information" is a technology company focusing on providing enterprise - level Agent intelligent body platforms and intelligent automation solutions for state - owned and central enterprises. Based on a cloud - native architecture, the company integrates AI, RPA, low - code, and process mining technologies, forming a product matrix centered around the enterprise - level intelligent body bit - Agent, covering multiple fields such as finance, energy, manufacturing, government, and military industry. Currently, its platform has been verified in complex scenarios of hundreds of large state - owned and central enterprises, with a penetration rate of over 30% in the central enterprise market.
(Image source/Enterprise)
Currently, the enterprise automation field is undergoing profound changes. As large - model technology promotes AI intelligent bodies into the stage of large - scale application, the limitations of traditional RPA, which is based on rule execution and cannot handle non - standard scenarios, are becoming increasingly prominent. In the core businesses of state - owned and central enterprises in energy, finance, manufacturing, etc., there are a large number of processes that are repetitive but have dynamically changing execution logics, such as operation and maintenance inspections, insurance underwriting, and procurement approvals. Traditional automation solutions need to preset all exception - handling rules, resulting in high development costs, frequent maintenance, and lack of flexibility. This pain point makes a new generation of intelligent automation tools with autonomous perception and decision - making capabilities a market necessity.
Yingke has learned that in response to this gap, "Jiuke Information" achieved a key breakthrough from "process automation" to "decision automation" this year. Its core product, bit - Agent, constructs a unique architecture of "separation of cognition and operation" by integrating multi - modal large models and RPA execution capabilities. This architecture does not rely on a single model and can support mainstream large models at home and abroad, such as DeepSeek, Tongyi Qianwen, and GPT - 4o. Through semantic understanding and operation abstraction of interface elements, it effectively avoids execution deviations caused by large - model hallucinations. In practical applications, bit - Agent can autonomously handle unpreset interaction changes, such as automatically identifying and signing new agreements in tax declarations and adjusting inspection routes according to real - time alarms in security operation and maintenance, greatly reducing the cost of manual intervention.
Image provided by the enterprise
In terms of product evolution, "Jiuke Information" further strengthened the collaborative ability of intelligent bodies in complex organizations. Through the intelligent organization management and control platform bit - Crew, enterprises can dispatch multiple intelligent bodies and RPA robots to form a "digital employee team" to achieve end - to - end business process automation across systems and departments. In a project with a leading automobile enterprise, this solution reduced the development cost of safety inspection templates from 30 person - days to 1 person - day, increasing efficiency by over 10 times and reducing the error rate from 15% to less than 1%. In addition, the bit - IDP intelligent document processing platform and bit - Bot intelligent knowledge assistant launched by the company last year further expanded its ability boundaries in unstructured data processing and private knowledge retrieval.
Wan Zhengyong, the CEO of "Jiuke Information", told Yingke that in terms of the competitive landscape, the domestic intelligent automation track is still in its early stage. Some traditional RPA manufacturers are limited by their technical paths and have difficulty getting rid of strong dependence on specific models; while some manufacturers entering from the large - model end encounter resistance in product implementation and large - scale application due to a lack of understanding of enterprise business processes and interface operations. The core product bit - Agent of "Jiuke Information" supports the embedding of enterprise private knowledge bases and a process solidification mechanism. The former allows the import of private - domain knowledge such as business rules and product manuals to guide automated decisions, and the latter can solidify verified task execution logics to reduce the uncertainty caused by large - model hallucinations. These two capabilities not only ensure the controllability and security of task execution but also meet the requirements of domestic deployment in the Xinchuang environment, forming a differentiated barrier.
(Image source/Enterprise)
In terms of market expansion, "Jiuke Information" has transformed from a single - tool provider to an industry - solution provider. The company has recently implemented multiple overall solutions in the fields of automobiles, finance, and energy. For example, it provides full - link operation and maintenance automation from safety inspection to exception handling for automobile enterprises and builds a full - link risk - control process automation system for financial institutions. In terms of Xinchuang replacement, in a project with Industrial Bank, the company has completed the domestic migration of over 1000 robots for more than 2000 existing processes under the original UiPath platform, covering over 800 business scenarios, and the return on investment in the first year of the project exceeded 1300%. Currently, the company's customer repurchase rate exceeds 80%, and the average customer unit price is over one million RMB.
(Image source/Enterprise)
In terms of the team, the core members of "Jiuke Information" come from companies such as CICC, Oracle, Huawei, Baidu, and Tencent, with over 20 years of experience in serving ultra - large enterprises. Wan Zhengyong, the founder of "Jiuke Information", graduated from Peking University and once served as the senior director of China Merchants Finance Technology and the head of the overseas R & D center of CICC, with rich experience in enterprise - level system R & D and team management; Fu Kai, the product VP, has led the implementation of multiple commercial products worth hundreds of millions of yuan. The low - code platform he built achieved creative revenue of over ten million yuan from a single customer at China Construction Bank Financial Technology; Dr. Deng Yan, the chief algorithm scientist, once served as a scientist in Google's AI team. The visual models and reinforcement learning algorithms he developed have given bit - Agent stronger interface understanding and response capabilities.
It is reported that after the completion of this nearly 100 - million - yuan Series B2 financing round, Jiuke Information has officially launched preparations for an IPO.
Views of the investor:
The investment head of Tejianfa Asia Merchant Fund said: As a leading intelligent automation enterprise incubated from China Merchants Group, Jiuke Information's product matrix built on a fully cloud - native architecture precisely addresses the core needs of state - owned and central enterprises' digital transformation. Its technical path of integrating "AI + RPA+ low - code + process mining" not only breaks the limitations of traditional automation but also achieves innovative breakthroughs in the track with the enterprise - level intelligent body product bit - Agent, forming a differentiated advantage in scenarios such as enterprise intelligent office. The company has been verified by hundreds of state - owned and central enterprises, with a penetration rate of over 30% in central enterprises. The high repurchase rate and customer unit price confirm market recognition. We are optimistic about its growth potential in the wave of Xinchuang and believe that this round of financing will help it deepen technology R & D and market expansion and continue to lead the intelligent automation upgrade of state - owned and central enterprises.