On September 22, the 2026 Shenzhou Holdings Ecosystem Partner Conference was held in Guangzhou under the theme 'Jieban Zhiyuan · Intelligence Must Act.' At the conference, Kejie Logistics, a subsidiary of Shenzhou Holdings, officially launched a new generation of enterprise-level supply chain intelligent system 'Jie Xiaoguo,' positioned as 'an AI intelligent agent that understands the supply chain better.' It features a simple three-step operation: 'open, input requests, receive results,' and, with an affordable price of 58 yuan per person per month, digitally replicates the experience of senior supply chain employees across every position.

Not being a 'general chat assistant', turning veteran experience into digital employees

During the on-site release, Zhang Hupo, CTO of Shenzhou Holdings Kejie, first raised a common pain point in the supply chain industry: behind more than 100 warehouses and nearly ten thousand operational issues every day, single systems like WMS and TMS can only answer fixed questions within their own domains. The ones who can truly make decisions and provide answers are the 'veteran experts' who have memorized all the rules. What Jiexiao Guo aims to do is to digitize and replicate these 'most skilled veteran experts'.

Unlike general AI platforms on the market that focus on creating PPTs, writing reports, or reading emails, Jie Xiao Guo targets specific and critical tasks in the supply chain—evaluating the effectiveness of marketing campaigns, reconciling inventory differences, organizing full truckload orders, repeatedly adjusting profit and loss, and verifying billing transactions. It employs a three-layer architecture of 'personality + skills + tools': clarifying the professional background and thinking style of an experienced employee (personality), equipping with job-specific skills such as managing storage locations and selecting routes (skills), and connecting to business systems like ERP (tools).

Over the past four months, the team has traversed 118 positions at Kuehne + Nagel Logistics and sorted out 115 requirements, resulting in more than 100 ready-to-use supply chain skills, covering eight major business areas including operational data dashboards, warehouse and distribution demand forecasting, logistics expectation management, returns reverse inspection, contract compliance review, bidding document preparation, profit and loss analysis dashboards, and more. When employees use it, they only need to directly 'throw' meeting links, documents, spreadsheets, or web pages at it, and it can independently break down tasks and read files, delivering summaries, dashboards, or ready-to-use finished files within minutes, with almost zero learning curve.

"Many platforms claim they can integrate experience, but only KJ Logistics, which has over twenty years in supply chain management, knows how to properly manage warehouses, organize storage locations, and reconcile accounts," Zhang Hupo said. "Only companies that truly understand the industry can provide knowledgeable intelligent agents."

A bill goes from 20 hours to 15 minutes, 58 yuan/month to drive organizational-level AI transformation

Whether a product is good or not ultimately depends on real business validation. Dong Shuai, General Manager of the Finance Department at Shenzhou Holdings, shared on-site that in the past, reconciliation scenarios required more than 50 people each month, involved more than 30 major billing rules, millions of orders, and over 30 invoices, with an average manual reconciliation cycle of up to 16 days; after introducing Jie Xiaoguo, overall process efficiency improved by 60%, and just the action of 'generating an invoice' was reduced from an average of 20 hours per person to 15 minutes, with each order traceable back to the calculation rules, while important judgments are still overseen by humans. Xia Xu, Deputy General Manager of the Marketing Department at Shenzhou Holdings Kejie, shared how he leveraged Jie Xiaoguo to achieve a capability leap from salesperson to super salesperson and then to frontline deployment engineer. He stated that the most valuable aspect of Jie Xiaoguo is its ability to solidify the experience and know-how of veteran staff into reusable skill packages, lowering the threshold for AI adoption in enterprises and expanding the capability boundaries of ordinary employees.

The earliest co-created client, Huazhi Liquor, verified the implementation speed of this model. Fu Zhenying, Director of the Information Center at Huazhi Liquor, explained that the collaboration involved 'tool platform + on-site FDE + business team,' where FDEs (Frontier Deployment Engineers) were stationed on-site to work with the business team to deconstruct rules and adjust scenarios. In just two days, they completed skill building and effect verification for the financial fund positions: reporting tasks were reduced from 2–3 days to 2–4 hours, transaction consolidation and sorting was shortened from a day and a half to 2 hours, and monthly asset inventory efficiency increased by more than 90%. Fu Zhenying stated that the real change was not how much AI accelerated processes, but how the job work model changed—people were freed from repetitive checks and shifted toward confirmation and judgment.

Also worth noting are the changes in delivery and business model. Jie Xiaoguo not only 'provides tools,' but accompanies customers through the entire process from cognition change, tool usage to process restructuring through the FDE model; the pricing is only 58 yuan per person per month, about 1,200 yuan per month for a 20-person team. During the internal promotion phase, the monthly active users have already exceeded 300.

At a time when AI is rapidly penetrating various industries, the supply chain has long been regarded as a 'tough nut to crack,' characterized by complex data, fragmented rules, and multiple systems. The launch of Jie Xiaoguo signifies that the competition among enterprise-level intelligent agents is shifting from 'who can chat better' to 'who understands the industry better.' Job skills accumulated over more than twenty years of industry know-how, a data foundation truly connected to business systems, combined with extremely low barriers to entry and cost, may be the key piece of the puzzle for AI in the supply chain to move from 'early trying' to large-scale implementation.

(Note: This article is commercial information published on CNR.cn. The content does not represent the opinions of this website and is for reference only.)

Follow exciting content