Content
# Intelligent Customer Service Using Spring AI Alibaba
> Spring AI Alibaba Repo: https://github.com/alibaba/spring-ai-alibaba
>
> Spring AI Alibaba Website: https://java2ai.com
>
> Spring AI Alibaba Website Repo: https://github.com/springaialibaba/spring-ai-alibaba-website
EcommSpringBot
EcommSpringBot is an enterprise-level intelligent customer service assistant system based on SpringAI Alibaba. It supports users to complete order inquiries, automatic cancellations, rule Q&A and other operations through natural language interaction. It integrates microservice architecture, RAG technology, Redis chat context memory and Elasticsearch vector database to realize a privatized and scalable intelligent agent service platform.
🌟 Project Highlights
• 🧠 Implemented enterprise intelligent agent dialogue capabilities based on SpringAI Alibaba (supports Tongyi Qianwen/OpenAI)
• 🔗 Modular design: Order service, middle-end service, SSE connection to MCP Server, and RAG knowledge base are clearly separated
• 🧾 Supports natural language query orders and automatic cancellation of orders
• 📚 Use Elasticsearch to build an enterprise knowledge base, supporting RAG retrieval enhancement
• 🧠 Use Redis to implement context memory function, supporting multi-turn conversations
🏗️ Module Structure
EcommSpringBot/
├── mall-order/ # Core order service (Spring Boot + MySQL)
├── mall-order-cmp-server/ # CMP middle-end service (encapsulates order capabilities)
├── mall-order-cmp_sso-client/ # Provides Controller interface + Redis chat context support
├── mall-order-es-rag/ # Vector knowledge base module (return and exchange rule documents → ES vector)
├── mall-order-graph-server/ # MCP graph computing service (based on Spring AI Alibaba)
└── README.md
📌 Features
✅ 1. Natural Language Operation Order Service
• Users can use natural language dialogues, such as:
• "I want to query the order of user USER1005"
• "Help me cancel the order of user USER1005"
Automatically match the interface through Spring Ai Alibaba Function Calling and call the mall-order service to complete.
⸻
✅ 2. SSE Real-time Reply + Redis Context Memory
• Use SSE (Server-Sent Events) to establish a streaming connection
• Implement user dialogue context memory based on Redis
• Support continuous context query and concise and natural human-computer interaction experience
⸻
✅ 3. Vector Retrieval Enhancement (RAG) + ES
• Organize the contents of return and exchange rules, invoice instructions, etc. to build a knowledge base and provide dynamic modification of the knowledge base content
• Use the text-embedding-v1 model to generate vectors and store them in Elasticsearch
• Users can automatically semantically match the content of the document and answer questions such as "How long is the return time for my product?"
🚀 Quick Start
Prerequisites
• JDK 17+
• Maven 3.9+
• Redis
• MySQL
• Elasticsearch 8.11+
Startup sequence
# Start the mall-order service
cd mall-order && mvn spring-boot:run
# Start mall-order-cmp-server
cd ../mall-order-cmp-server && mvn spring-boot:run
# Start the sso-client interface layer (including AI intelligent agent interface)
cd ../mall-order-cmp_sso-client && mvn spring-boot:run
# Start the es-rag service (for knowledge base management)
cd ../mall-order-es-rag && mvn spring-boot:run
📖
📄 LICENSE
This project is open sourced based on Apache 2.0 License and can be used, modified and distributed freely.
⸻
If you think this project is valuable, you are welcome to Star⭐️ or Fork🍴!
📬 Contact the author: 996766130@qq.com
## Front-end page

Connection Info
You Might Also Like
AP2
AP2 provides code samples and demos for the Agent Payments Protocol.
nuwax
Nuwax AI enables easy building and deployment of private Agentic AI solutions.
daydreams
Daydreams is an AI agent framework in TypeScript for scalable and composable...
concierge
Concierge is a platform for community engagement and scheduling demos.
google-meta-ads-ga4-mcp
MCP server for Google Ads, Meta Ads & GA4 — works with ChatGPT, Claude,...
mcp-server-airbnb
A Desktop Extension for advanced Airbnb search and listings with detailed filtering.