Xingyifu and Huilaidian Unveil Modou Cloud AI Ecosystem at 2026 Yunqi Conference

The Yunqi Conference is Alibaba Group's annual event, presented by Alibaba Cloud with partners and global customers. This year's theme was Intelligence for Practical Use, with Agentic AI as the core focus linking chips, cloud infrastructure, model capabilities and model services, and Agentic applications. The conference centered on changes in production methods.

During the Financial Intelligent Agent in Practice Summit, held under the theme Trusted AI in Action, participants from finance and technology discussed how financial agents can become executable and trustworthy. Chen Yi, chairman of Fujian Xingyi Payment Technology Co., Ltd., and general manager Lin Jian were invited to the summit, joining representatives from wealth management, robo-advisory and insurtech to share practices for Agentic financial products.

Lin Jian focused on merchant operations and presented the Luban Super AI Agent, described as an industry first. Luban uses an OpenClaw plus Hermes dual-engine architecture, combines a self-developed Channel Server strategy with ecosystem capabilities from Alibaba Cloud, Dianjin and Alipay, and has developed more than 36 core operational capabilities to support intelligent merchant services that are tailored to each merchant and each agent. It supports out-of-the-box use and can be connected to instant messaging tools including WeChat, WeCom and DingTalk, making conversation-as-operation a new form of human-machine collaboration.

For global merchant operations, Luban has built three ecosystem capabilities. A PC ecosystem based on a deep workbench supports one-click access to external systems and interconnection with agents. A hardware ecosystem covering cashier, sensing and smart terminal IoT devices brings multi-source data into the operations engine. An Agent Payment service builds a global trusted payment hub for agents, connecting merchants, acquirers and card organizations and linking demand understanding with audit traceability across the payment chain. Luban currently serves hundreds of thousands of merchants and handles more than 35 billion token calls per day, entering business processes as a commercial example of an executable and trustworthy financial agent.

At the conference, Xingyifu and Huilaidian also exhibited the self-developed Modou Cloud AI application platform. The platform relies on business intelligence applications, enterprise-level solutions and basic ecosystem applications to break down barriers between AI capabilities and industry scenarios. Business intelligence applications include Luban for global merchant operations and Moke Studio, a one-stop AI video generation and publishing platform. Enterprise-level solutions include AI marketing covering the full merchant operation chain, an AI merchant review system that improves review efficiency, and Agent Payment services that build a global commercial transaction and trusted payment hub for agents. Basic ecosystem applications include Modou Market, an AI application ecosystem service platform, and modorouter, which provides unified access to global AI model capabilities.

Based on AI-native capabilities for unified connection, invocation, orchestration and billing, Modou Cloud is pushing AI implementation at scale across industries. It currently provides AI capabilities to financial enterprises, technology companies, small and medium-sized businesses and individual creators. AI marketing has handled more than 60,000 marketing tasks and generated nearly 30,000 marketing materials. The AI merchant review system has replaced 75 percent of manual workload, with cumulative review items reaching 120 million and more than 5 million merchants reviewed.

On the exhibition stage, the companies gave in-depth presentations on Luban and Moke Studio. Luban uses self-evolution and planning capabilities to extract and analyze merchant information, customize operation plans through modular skills and a tool ecosystem, and improve with long-term memory and accumulated data. Moke Studio has an independently developed Agent intelligent scheduling engine that coordinates mainstream large models through multi-agent collaboration, connects the full chain from natural language to deployable video, and includes compliance review, allowing one-time generation to meet expectations.

As AI moves from a technology boom into deeper industry use, the key competition is shifting from model parameters to implementation efficiency. Xingyifu and Huilaidian said they will continue to deepen AI ecosystem development and use more open ecosystem collaboration to promote the implementation and iteration of AI capabilities in industry scenarios, advance digital commerce and make payments more valuable and opening stores simpler.