Content
# EeveeAgent
## Project Introduction
EeveeAgent is an enterprise-level Agentic RAG platform that covers the entire chain from front-end construction to Java-based AI back-end engineering applications, including knowledge base document upload, intelligent Q&A, and more.
- **Knowledge Base Management**: Multi-format document upload, parsing, pre-processing, chunking, enhancement, and vectorization.
- **Intent Recognition**: Tree-like multi-level classification, precise intent positioning, dedicated intent recognition prompts, TopK, and other core parameters.
- **Multi-channel Retrieval**: Multi-channel strategy parallel recall, post-processing responsibility chain, deduplication, and reordering for precise retrieval.
- **MCP Integration**: Non-knowledge-based intent automatic parameter invocation of business tools, seamless integration of retrieval and tools.
- **Front-end Interface**: Modern Vue 3 front-end application.
## Project Core Design
The project adopts a front-end and back-end separation architecture, with the back-end divided into four Maven modules based on responsibilities.
The `core-framework` layer provides general capabilities unrelated to business, the `eevee-ai` layer shields differences between different model suppliers, and the `main-application` layer focuses on business logic. Switching model suppliers does not require changing business code, and switching business logic does not affect infrastructure.
## Console Display
EeveeAgent provides a complete visual console that covers two types of user scenarios: ordinary users and administrator users. The interface is concise, intuitive, and efficient.
The system has undergone multiple rounds of AI-assisted design optimization to ensure functional integrity while providing a more modern and friendly interactive experience.
### 1. User Q&A Interface
Users can input questions directly in the input box on the EeveeAgent homepage to initiate Q&A. The system also supports enabling **Deep Thinking Mode** for higher-quality answers.
- Supports natural language input
- Supports quick filling of example questions
- Supports deep thinking mode
### 2. Management Background
EeveeAgent provides a fully functional management background for system configuration and operation management. Administrators can complete model management, system configuration, and data management operations through the background.
## Launch EeveeAgent
## Environmental Preparation
### 1. IDE Installation
The project requires IntelliJ IDEA 2023 or higher. If you don't have a professional version, you can use the community free version.
### 2. JDK Installation
The Ragent system framework is developed based on SpringBoot3, which requires JDK version 17 or higher.
## Project Clone and Configuration
### 1. Clone Project
Open the GitHub project address and copy the SSH clone address.
### 2. Configure Maven Environment
You can use IDEA's default Maven to compile the project. If compilation fails, try changing the Maven version.
### 3. Configure JDK17
After cloning the project, open `Project Structure...` configuration in IDEA and check the JDK version.
## Project Update
The project may continue to iterate and fix issues. It is recommended to pull the latest code before starting development.
## Docker Middleware Deployment
EeveeAgent AI project uses Docker to install middleware. If your computer does not have Docker installed, please complete the installation first.
### 1. MySQL
```bash
docker run -d \
--name eevee-mysql \
-p 3306:3306 \
-e MYSQL_ROOT_PASSWORD=123456 \
-e MYSQL_DATABASE=eeveeagent \
-e MYSQL_USER=admin \
-e MYSQL_PASSWORD=123456 \
-e TZ=Asia/Shanghai \
-v mysql-data:/var/lib/mysql \
mysql:8.0 \
--character-set-server=utf8mb4 \
--collation-server=utf8mb4_unicode_ci
```
### 2. Redis Installation
```bash
docker run -d \
--name eevee-redis \
-p 6379:6379 \
-e REDIS_PASSWORD=123456 \
-v redis-data:/data \
redis:7.2-alpine \
redis-server --requirepass 123456 --appendonly yes
```
### 3. Milvus Vector Database Stack
If Docker Compose is not installed, please install it first.
#### 3.1 Create Working Directory
```bash
mkdir -p ~/milvus && cd ~/milvus
```
#### 3.2 Download Milvus Configuration File
```bash
curl -O https://raw.githubusercontent.com/milvus-io/milvus/
v2.6.6/deployments/docker/standalone/docker-compose.yml
```
#### 3.3 Start Milvus
```bash
docker-compose up -d
```
#### 3.4 Check Startup Status
```bash
docker-compose ps
```
## Client Tools
### 1. Redis Desktop
Use [AnotherRedisDesktopManager](https://github.com/qishibo/AnotherRedisDesktopManager/releases) as the Redis client connection tool.
### 2. Attu Management Interface
This is the visual interface for the Milvus database.
### 3. RustFS
RustFS provides a web console for managing object storage, creating buckets, uploading and downloading files, and configuring access policies.
## IDEA Integrate Docker
You can start all middleware with one click using the IDEA service.
### 1. Pre-preparation
Enable Docker remote API (essential for IDEA connection).
### 2. IDEA Configure Docker Connection
1. Open IDEA → `File` → `Settings` (shortcut key `Ctrl+Alt+S`)
2. Find `Build, Execution, Deployment` → `Docker`
3. Click the **+** button in the upper right corner to create a new Docker configuration
4. Select TCP socket, fill in the address: `tcp://your IP address:2375`
5. Click **Test connection**, display **Connection successful** to connect successfully
## MySQL Database Initialization
## SQL Script File Description
In the `/resources/database` directory of the project, there are two core SQL script files:
| File Name | Function Description | Execution Order |
| :-------- | :------------------- | :-------------- |
| schema.sql | Contains database creation statements, table structure definitions, and index configurations | 1 |
| init_data.sql | Contains business initialization data, such as system configuration, role permissions, etc. | 2 |
## Contribution Guide
Welcome to contribute code and suggestions!
1. **Fork** project repository
2. **Create** functional branch
3. **Submit** code changes
4. **Push** to remote branch
5. **Create** Pull Request
## License
This project uses the MIT license. See the LICENSE file for details.
**EeveeAgent** - Intelligent Agent System
