On June 16, 2026, the 2026 Cross-border E-commerce Development Conference, co-sponsored by the China Association for Trade in Services and the World Trade Network Alliance, was grandly held in Langfang, Hebei.
With the theme of “Building a Trustworthy Trade Mechanism to Help Brands Go Global”, the conference attracted people from universities, enterprises, associations and other walks of life to gather together to discuss the future development trends and innovation paths of the cross-border e-commerce industry.
At the meeting, Zhang Hupo, chief technology officer of Kejie of Digital China Holdings Co., Ltd. (referred to as “China Holdings”), delivered a keynote speech, detailing the innovative application and core advantages of China Holdings’ “supply chain AI control tower” solution in cross-border e-commerce scenarios. This speech attracted widespread attention from the guests attending the meeting.
At the same time, the conference released the “2025 China Cross-border E-commerce Annual Development Report”.
In the “2025 China Cross-border E-commerce Service Providers Comprehensive Competitiveness List”, China Holdings (Kejie) successfully entered the list, demonstrating its deep strength and industry recognition in the field of cross-border e-commerce services.
The report specifically mentioned that the supply chain intelligence “Xiaojin” self-developed by Kejie Logistics, a subsidiary of China Holdings, performed well in the actual Double Eleven event in 2025 and became a typical case of AI technology empowering the supply chain.
Zhang Hupo pointed out in his speech that the current cross-border e-commerce industry generally faces the dilemma of AI technology being difficult to truly integrate into the business decision-making process.
Although most companies have deployed data visualization large screens, AI chat robots and other tools, the company’s data has not been fully utilized, and the WMS data in the backend of each platform and overseas warehouses is idle. In terms of processes, overseas warehouse systems, platform backends, and first-leg logistics systems are not connected to each other, resulting in AI prediction results unable to affect replenishment and allocation decisions.
In response to these problems, China Holdings launched the “Supply Chain AI Control Tower”.
The supply chain AI control tower adopts a three-layer architecture design.
The bottom layer connects overseas warehouse WMS, ERP, TMS, as well as various system data such as the backend of major e-commerce platforms and exchange rate and tariff data sources; the middle-level AI workbench for each position uses the “Xiaojin” intelligent agent to provide decision-making support for the six core nodes of demand forecasting, head-trip planning, overseas warehouse deployment, multi-platform replenishment, performance warning, and business attribution; the top-level AI control tower provides full-link visibility, collaborative decision-making, and attribution feedback capabilities, realizing the leap from “seeing data” to “assisted decision-making.”
For example, in the demand forecasting process, AI can predict the sales volume of each SKU and each platform in the peak season, and provide stocking suggestions for different platforms and warehouses; in the first-leg planning process, AI optimizes sea and air transportation schedules and costs; in the overseas warehouse allocation process, AI generates cross-warehouse allocation plans through multi-warehouse inventory thermal analysis.
case; in the multi-platform replenishment process, AI formulates differentiated replenishment strategies and automatically triggers replenishment instructions; in the performance warning process, AI identifies out-of-stock, overage, and return abnormalities in real time and provides disposal plans; in the operation attribution process, AI breaks down the specific impact of exchange rates, logistics, and platform fees on gross profit.
The “2025 China Cross-border E-commerce Annual Development Report” believes that the role of China’s cross-border e-commerce service providers is gradually moving from a simple back-end support force in the early days to front-end co-creation.
In the extreme scenario of peak orders of 110 million on Double 11 in 2025, the peak order volume per second of the “Xiaojin” agent increased by 280% year-on-year.
In daily operations, data query efficiency is increased by 90%, customer service workload is reduced by 50%, complex document picking and sorting efficiency is increased by 20%, packaging efficiency is increased by 15%, and the efficiency of order production positions is increased by 3 times.
At the same time, it can also participate in the formulation of corporate inventory strategies and provide customers with decision-making support such as warehouse settlement recommendations and new warehouse location plans. This shows that logistics service providers are forming a new type of in-depth co-creation partnership with manufacturing companies.
In the cross-border e-commerce scenario, industry characteristics such as long shipping cycles, different inventory rules on multiple platforms, dense peak season points, and exchange rate fluctuations affecting gross profits make the effective use of AI value a key factor in determining the outcome of competition.
China Holdings Kejie’s supply chain AI control tower will provide solid intelligent support for Chinese brands to go overseas, helping Chinese brands to develop steadily in the global market.
