AI_CRM
智能对话式客户关系管理系统 / Intelligent Conversational Customer Relationship Management System
快速导航 | Quick Nav: 🇨🇳 中文 | 🇺🇸 English
🇨🇳 中文版 | Chinese
🚀 立即体验
| 体验方式 | 地址/账号 | 说明 |
|---|---|---|
| 🌐 在线演示站 | https://www.72crm.com/wkaicrm | 一键访问,推荐首选 |
| 🔑 演示站账号 | 点击“免费体验”直接注册体验 | 用于登录在线演示站 |
| 💬 帮助与讨论 | 前往社区论坛 | 反馈问题、交流想法 |
提示:在线演示站已预置示例数据和客户信息,您可以直接登录并体验所有核心功能。
✨ 它能做什么?
| 功能模块 | 核心价值 |
|---|---|
| 💬 AI 对话助手 | 像同事一样询问业务:“上一季度华东区的销售冠军是谁?”,系统可结合结构化数据与知识库文档,生成智能回答。 |
| 🧠 知识库 RAG 增强 | 赋予 AI“记忆”:上传公司产品手册、合同、会议纪要,AI 助手能基于这些文档内容进行精准问答和总结。 |
| 👥 智能客户管理 | 一体化客户视图:集中管理客户信息、联系人、跟进记录,并通过 AI 自动分析客户阶段与需求。 |
| ✅ AI 任务生成 | 自动创建工作项:在对话或分析客户后,可指令 AI 创建待办任务,并自动设置优先级与提醒。 |
| 🔗 无缝团队协作 | 信息实时同步:客户动态、任务分配、知识更新均在团队内即时同步,促进高效协作。 |
🛠️ 技术栈
- 后端: Java 21 + Spring Boot 3.x + Spring AI + PostgreSQL + Redis + MinIO
- 前端: Vue 3 + TypeScript + Element Plus + Tailwind CSS
- 部署: 支持 Docker Compose 一键部署,提供完整生产环境配置。
后端技术栈明细
| 技术 | 版本 | 说明 |
|---|---|---|
| Java | 21 | 编程语言 |
| Spring Boot | 3.3.12 | 应用框架 |
| Spring AI | 1.0.0 | AI/LLM 集成 (支持 OpenAI 兼容 API) |
| PostgreSQL | 17 | 主数据库 |
| MyBatis-Plus | 3.5.7 | 数据持久层框架 |
| Redis | - | 缓存与会话管理 |
| MinIO | - | 对象存储(用于文档、文件) |
前端技术栈明细
| 技术 | 版本 | 说明 |
|---|---|---|
| Vue | 3.4 | 前端框架 |
| TypeScript | 5.5 | 类型安全 |
| Element Plus | 2.8 | UI 组件库 |
| Pinia | 2.2 | 状态管理 |
| Tailwind CSS | 3.4 | 实用 CSS 框架 |
| Vite | 5.4 | 构建工具 |
📁 项目结构
wk_ai_crm/
├── backend/ # 后端 Spring Boot 项目
│ ├── src/main/java/ # Java 源码
│ ├── src/main/resources/ # 配置文件
│ └── pom.xml # Maven 配置
├── frontend/ # 前端 Vue 项目
│ ├── src/ # 前端源码
│ └── package.json # npm 配置
└── docker/ # Docker 部署配置
├── docker-compose.yaml # 编排文件
└── nginx/ # Nginx 配置
└── LICENSE.md # 协议文件
└── README.md # 本文档
⚡️ 快速开始(本地开发)
如果你想在本地运行或进行二次开发,请遵循以下步骤。
先决条件
- JDK 21+, Node.js 18+, Maven 3.8+
- PostgreSQL 17, Redis 6+
- (可选) Docker & Docker Compose
1. 克隆项目
git clone https://github.com/WuKongOpenSource/AI_CRM.git
cd AI_CRM
2. 后端启动
cd backend
mvn clean install
mvn spring-boot:run
# API 服务将在 http://localhost:8088 运行
# API 文档 (Knife4j): http://localhost:8088/doc.html
3. 前端启动
cd frontend
npm install
npm run dev
# 前端将在 http://localhost:5173 运行
4. 使用 Docker 一键部署(推荐)
cd docker
docker-compose up -d
# 访问 http://localhost 即可 (Nginx 反向代理了前后端)
配置文件:首次运行前,请根据 backend/src/main/resources/application.yml 中的注释,配置数据库、AI API 密钥(如 OpenAI、DeepSeek 等)等必要信息。
配置说明
主要配置文件:backend/src/main/resources/application.yml
数据库配置
spring:
datasource:
url: jdbc:postgresql://localhost:5432/wk_ai_crm
username: postgres
password: your_password
Redis 配置
spring:
data:
redis:
host: localhost
port: 6379
password: your_password
database: 7
AI 服务配置
spring:
ai:
openai:
api-key: your_api_key
base-url: https://api.openai.com/v1/ # 或其他兼容 API
chat:
options:
model: gpt-4
MinIO 对象存储配置
minio:
enabled: true
endpoint: http://localhost:9000
access-key: minioadmin
secret-key: minioadmin
bucket: ai-crm
WeKnora 知识库服务配置
weknora:
enabled: true
base-url: http://localhost:8080/api/v1
api-key: your_api_key
knowledge-base-id: your_kb_id
API 文档
启动后端服务后,访问 Knife4j API 文档:
http://localhost:8088/doc.html
默认账号
- 用户名:
admin - 密码:
123456a
模型配置
-安装完成需要到“系统设置”的“API/AI”中进行 AI 大模型配置,输入对应的 key,否则对话会出错。
🤝 欢迎贡献
AI CRM 正处于快速成长阶段,我们热烈欢迎任何形式的贡献!
- 🐛 报告问题:使用 GitHub Issues 提交 Bug 或新功能建议。
- 🔧 提交代码:请阅读我们的贡献指南(待创建),了解开发流程和代码规范。
- 📖 完善文档:帮助改进文档、翻译,让项目更易懂。
- 💡 分享想法:在社区论坛分享你的使用场景或优化建议。
📄 许可证
本项目基于 MIT License 开源。这意味着你可以自由地使用、修改和分发代码,但需保留原作者的版权声明。
❓ 常见问题
Q:AI 模型支持哪些? A:默认支持任何提供 OpenAI 兼容 API 的模型(如 OpenAI GPT 系列、DeepSeek、Ollama 本地模型等)。在后台“系统设置”的“API/AI”配置中填入对应 API Key 即可。
Q:商业使用时数据安全吗? A:项目可完全私有化部署,所有数据(客户、文档、AI 交互)均保存在您自己的服务器中,确保数据安全。
Q:如何获取更多帮助? A:您可以访问项目的 社区论坛 提问或搜索现有答案。
🇺🇸 English Version | 英文版
🚀 Try It Now
We strongly recommend you first experience the power of AI CRM through the following methods.
| Experience | Address/Account | Notes |
|---|---|---|
| 🌐 Live Demo | https://www.72crm.com/wkaicrm | One-click access, recommended |
| 🔑 Demo Account | Click "Free Trial" to register and experience directly | For logging into the live demo site |
| 💬 Help & Discussion | Community Forum | Report issues and share ideas |
Tip: The live demo comes pre-loaded with sample data and customer information. You can log in directly and experience all core features.
✨ What Can It Do?
AI CRM is more than a traditional CRM; it‘s an AI partner that understands your business.
| Feature | Core Value |
|---|---|
| 💬 AI Conversational Assistant | Ask about business like talking to a colleague: “Who was the sales champion in East China last quarter?” The system can generate intelligent answers by combining structured data and knowledge base documents. |
| 🧠 Knowledge Base RAG Enhancement | Give AI “memory”: Upload company product manuals, contracts, meeting minutes. The AI assistant can provide precise Q&A and summaries based on these documents. |
| 👥 Intelligent Customer Management | Unified customer view: Centrally manage customer information, contacts, follow-up records, with AI automatically analyzing customer stages and needs. |
| ✅ AI Task Generation | Automatically create work items: After conversations or customer analysis, instruct AI to create to-do tasks with automatic priority and reminders. |
| 🔗 Seamless Team Collaboration | Real-time information sync: Customer updates, task assignments, and knowledge updates are instantly synchronized within the team for efficient collaboration. |
🛠️ Technology Stack
This is a full-stack open-source project with a modern and stable technology stack.
- Backend: Java 21 + Spring Boot 3.x + Spring AI + PostgreSQL + Redis + MinIO
- Frontend: Vue 3 + TypeScript + Element Plus + Tailwind CSS
- Deployment: Supports one-click deployment via Docker Compose with complete production environment configuration.
Backend Tech Stack Details
| Technology | Version | Purpose |
|---|---|---|
| Java | 21 | Programming Language |
| Spring Boot | 3.3.12 | Application Framework |
| Spring AI | 1.0.0 | AI/LLM Integration (OpenAI-compatible API) |
| PostgreSQL | 17 | Primary Database |
| MyBatis-Plus | 3.5.7 | ORM Framework |
| Redis | - | Cache & Session Management |
| MinIO | - | Object Storage (for docs, files) |
Frontend Tech Stack Details
| Technology | Version | Purpose |
|---|---|---|
| Vue | 3.4 | Frontend Framework |
| TypeScript | 5.5 | Type Safety |
| Element Plus | 2.8 | UI Component Library |
| Pinia | 2.2 | State Management |
| Tailwind CSS | 3.4 | Utility-first CSS Framework |
| Vite | 5.4 | Build Tool |
📁 Project Structure
wk_ai_crm/
├── backend/ # Backend Spring Boot Project
│ ├── src/main/java/ # Java Source Code
│ ├── src/main/resources/ # Configuration Files
│ └── pom.xml # Maven Configuration
├── frontend/ # Frontend Vue Project
│ ├── src/ # Frontend Source Code
│ └── package.json # npm Configuration
├── docker/ # Docker Deployment Configuration
│ ├── docker-compose.yaml # Orchestration File
│ └── nginx/ # Nginx Configuration
├── LICENSE.md # License File
└── README.md # This Document
⚡️ Quick Start (Local Development)
If you want to run it locally or contribute to development, please follow these steps.
Prerequisites
- JDK 21+, Node.js 18+, Maven 3.8+
- PostgreSQL 17, Redis 6+
- (Optional) Docker & Docker Compose
1. Clone the Repository
git clone https://github.com/WuKongOpenSource/AI_CRM.git
cd AI_CRM
2. Start the Backend
cd backend
mvn clean install
mvn spring-boot:run
# API server will run at http://localhost:8088
# API Docs (Knife4j): http://localhost:8088/doc.html
3. Start the Frontend
cd frontend
npm install
npm run dev
# Frontend will run at http://localhost:5173
4. One-Click Deployment with Docker (Recommended)
cd docker
docker-compose up -d
# Visit http://localhost (Nginx reverse proxies frontend and backend)
Configuration: Before first run, configure necessary information like database and AI API keys (e.g., OpenAI, DeepSeek) according to comments in backend/src/main/resources/application.yml.
Configuration Guide
Main configuration file: backend/src/main/resources/application.yml
Database Configuration
spring:
datasource:
url: jdbc:postgresql://localhost:5432/wk_ai_crm
username: postgres
password: your_password
Redis Configuration
spring:
data:
redis:
host: localhost
port: 6379
password: your_password
database: 7
AI Service Configuration
spring:
ai:
openai:
api-key: your_api_key
base-url: https://api.openai.com/v1/ # Or other compatible API
chat:
options:
model: gpt-4
MinIO Object Storage Configuration
minio:
enabled: true
endpoint: http://localhost:9000
access-key: minioadmin
secret-key: minioadmin
bucket: ai-crm
WeKnora Knowledge Base Service Configuration
weknora:
enabled: true
base-url: http://localhost:8080/api/v1
api-key: your_api_key
knowledge-base-id: your_kb_id
API Documentation
After starting the backend service, access the Knife4j API documentation at:
http://localhost:8088/doc.html
Default Account
· Username: admin · Password: 123456a
Model Configuration
After installation, you must go to "System Settings" -> "API/AI" to configure the AI large model by entering the corresponding API key. Otherwise, the conversation feature will fail.
🤝 Welcome Contributions
AI CRM is in a rapid growth phase, and we warmly welcome contributions of all forms!
- 🐛 Report Issues: Use GitHub Issues to submit bugs or feature suggestions.
- 🔧 Submit Code: Pull Requests are welcome.
- 📖 Improve Documentation: Help with docs or translations.
- 💡 Share Ideas: Discuss in our Community Forum.
📄 License
This project is open source under the MIT License. This means you are free to use, modify, and distribute the code, provided that the original copyright notice is retained.
❓ FAQ
Q: Which AI models are supported? A: By default, it supports any model providing an OpenAI-compatible API (e.g., OpenAI GPT series, DeepSeek, Ollama local models). Configure the corresponding API Key in the backend “System Settings” -> “API/AI” section.
Q: Is data safe for commercial use? A: The project can be fully self-hosted. All data (customers, documents, AI interactions) is stored on your own servers, ensuring data security.
Q: How to get more help? A: You can visit the project’s Community Forum to ask questions or search for existing answers.
🌟 项目动态 / Project Updates
最新 / Latest: 项目预览版 v0.1.0 正式开源!/ Preview v0.1.0 officially open-sourced!
如果 AI CRM 对你有帮助,请给我们一个 ⭐️ Star!这是对我们开源工作的最大鼓励。
If AI CRM helps you, please give us a ⭐️ Star! It's the greatest encouragement for our open-source work.
