# rag-ai

**Repository Path**: 51tech/rag-ai

## Basic Information

- **Project Name**: rag-ai
- **Description**: 基于Spring-AI的RAG检索增强生成系统，支持文档资料库搜索的数据库自动查询两种模式。
1、底座使用ruoyi框架，springboot3和vue3版本
2、数据库使用MariaDB，支持向量搜索
3、嵌入模型使用ollama，模型可根据自己服务器性能选择，开发阶段使用deepseek-r1:7b做测试
- **Primary Language**: Java
- **License**: LGPL-3.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 2
- **Forks**: 3
- **Created**: 2026-02-06
- **Last Updated**: 2026-06-02

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# RAG-AI Intelligent Q&A and Retrieval-Augmented Generation System

## Project Overview

RAG-AI is an intelligent Q&A and Retrieval-Augmented Generation (RAG) system extended from the RuoYi framework. It integrates multiple AI capabilities to provide users with intelligent conversational interaction, data retrieval, and knowledge-based question answering services. The system adopts a frontend-backend separated architecture, supports WebSocket real-time communication, and features high efficiency and scalability.

## Main Features

### Intelligent Dialogue Engine
- **RAG Engine**: Knowledge-based Q&A powered by vector-based retrieval and enhanced generation
- **QA Engine**: Professional question-answering system supporting context understanding and multi-turn dialogue
- **Database Engine**: Natural language to SQL conversion for intelligent database querying
- **Streaming Response**: Supports SSE and WebSocket for real-time response streaming

### Knowledge Base Management
- **Document Management**: Supports uploading, training, and retrieving documents in multiple formats
- **Data Source Configuration**: Supports configuration of multiple database connections and table schemas
- **Field Mapping**: Flexible mapping between database fields and the Q&A system
- **Incremental Training**: Supports real-time updates and retraining of the knowledge base

### Dialogue Management
- **Conversation History**: Saves and retrieves historical conversation records
- **Context Memory**: Maintains coherence across multi-turn dialogues
- **Role Configuration**: Supports expert role configurations for different scenarios
- **Permission Control**: Fine-grained permission management based on the RuoYi framework

### System Monitoring
- **Online Users**: Real-time monitoring of user login status
- **Operation Logs**: Records user operation trails
- **Service Monitoring**: Monitors server resource utilization
- **Cache Management**: Monitors Redis cache status

## Technology Stack

### Backend Technologies
- **Core Framework**: Spring Boot 2.7.x
- **Security Framework**: Spring Security + JWT
- **Caching**: Redis
- **Database**: MySQL + Druid connection pool
- **AI Framework**: Spring AI
- **WebSocket**: Spring WebSocket
- **Task Scheduling**: Quartz
- **Code Generation**: MyBatis-Plus + Velocity

### Frontend Technologies
- **Framework**: Vue3 + Vite
- **State Management**: Pinia
- **UI Components**: Element Plus
- **Routing**: Vue Router
- **HTTP Client**: Axios
- **Rich Text Editor**: WangEditor

## Quick Start

### Environment Requirements
- JDK 1.8+
- MySQL 5.7+
- Redis 3.0+
- Node.js 16.0+

### Backend Deployment

1. **Create Database**
```sql
CREATE DATABASE rag_ai DEFAULT CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;
```

2. **Import SQL Script**
```bash
mysql -u root -p rag_ai < sql/java_ai.sql
```

3. **Modify Configuration Files**
```yaml
# application-druid.yml
spring:
  datasource:
    url: jdbc:mysql://localhost:3306/rag_ai
    username: your_username
    password: your_password
  redis:
    host: localhost
    port: 6379

# application.yml
ai:
  vector-similarity: 0.7
  vector-limit: 10
```

4. **Start Service**
```bash
cd /data/skyline/skyline-repo-cache/45/0a89bed8ed1f00581f3483237f778dc016708a891d7d30222e5afd2bbef9a7
./mvnw spring-boot:run
```

### Frontend Deployment

1. **Install Dependencies**
```bash
cd ui
npm install
```

2. **Start Development Server**
```bash
npm run dev
```

3. **Build Production Version**
```bash
npm run build:prod
```

## Access Addresses

- **Backend Service**: `http://localhost:8080`
- **Frontend Interface**: `http://localhost:80`
- **API Documentation**: `http://localhost:8080/swagger-ui.html`

## Directory Structure

```
rag-ai/
├── src/main/java/com/ruoyi/
│   ├── chat/                    # Chat-related modules
│   │   ├── common/              # Common components
│   │   ├── config/              # Configuration classes
│   │   ├── controller/          # Controllers
│   │   ├── domain/              # Entity classes
│   │   ├── engine/              # Chat engines
│   │   ├── response/            # Response handlers
│   │   ├── service/             # Service layer
│   │   └── ws/                  # WebSocket
│   ├── rag/                     # RAG core module
│   │   ├── advisor/             # AI advisor
│   │   ├── controller/          # Controllers
│   │   ├── domain/              # Entity classes
│   │   ├── engine/              # Engine implementations
│   │   ├── mapper/              # Data mappers
│   │   ├── service/             # Service layer
│   │   ├── tools/               # Utility classes
│   │   ├── utils/               # Utility classes
│   │   └── vo/                  # View objects
│   └── ...                      # Other modules
├── ui/                           # Frontend project
│   ├── src/
│   │   ├── api/                 # API interfaces
│   │   ├── views/               # View components
│   │   ├── store/               # State management
│   │   └── utils/               # Utility classes
│   └── ...
└── sql/                          # SQL scripts
```

## API Interfaces

### Chat APIs

| Endpoint | Method | Description |
|----------|--------|-------------|
| `/ai/chat/rag/{userId}` | GET | RAG Q&A streaming response |
| `/ai/chat/qa/{userId}` | GET | QA Q&A streaming response |
| `/ai/chat/db/{userId}` | GET | Database query streaming response |
| `/ai/chat/stream/{userId}` | GET | General chat streaming response |

### RAG Configuration APIs

| Endpoint | Method | Description |
|----------|--------|-------------|
| `/rag/config` | CRUD | AI configuration management |
| `/rag/document` | CRUD | Document management |
| `/rag/table` | CRUD | Table management |
| `/rag/field` | CRUD | Field management |
| `/rag/datasource` | GET | List of data sources |

## License

This project is open-sourced under the MIT License.