# pgvector

**Repository Path**: ease-meta/pgvector

## Basic Information

- **Project Name**: pgvector
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2026-04-21
- **Last Updated**: 2026-04-21

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# SpringAI PgVector Integration with Ollama

This project demonstrates how to integrate SpringAI with PgVector for vector database storage and use Ollama to connect to the local qwen3.5:27b model.

## Prerequisites

1. **PostgreSQL with PgVector extension**
   - Install PostgreSQL (version 14+ recommended)
   - Install PgVector extension: `CREATE EXTENSION vector;`
   - Create a database named `vector_db`

2. **Ollama**
   - Install Ollama from https://ollama.com/
   - Pull the qwen3.5:27b model: `ollama pull qwen3.5:27b`
   - Start the Ollama service

3. **Java 17+**
   - Install JDK 17 or higher

4. **Maven**
   - Install Maven 3.6+ (for building the project)

## Project Structure

```
src/
├── main/
│   ├── java/com/example/springai_pgvector/
│   │   ├── SpringaiPgvectorApplication.java   # Main application class
│   │   ├── config/VectorConfig.java           # Configuration for vector store
│   │   ├── controller/VectorController.java   # REST API endpoints
│   │   ├── entity/VectorEmbedding.java        # Vector embedding entity
│   │   ├── repository/VectorEmbeddingRepository.java  # JPA repository
│   │   └── service/VectorService.java         # Business logic for vector operations
│   └── resources/
│       └── application.properties            # Application configuration
└── test/
    └── java/com/example/springai_pgvector/
        └── SpringaiPgvectorApplicationTests.java  # Test class
```

## Configuration

The application.properties file contains the following configuration:

```properties
# PostgreSQL Configuration
spring.datasource.url=jdbc:postgresql://localhost:5432/vector_db
spring.datasource.username=postgres
spring.datasource.password=postgres
spring.jpa.hibernate.ddl-auto=update
spring.jpa.properties.hibernate.dialect=org.hibernate.dialect.PostgreSQLDialect

# Ollama Configuration
spring.ai.ollama.base-url=http://localhost:11434
spring.ai.ollama.chat.model=qwen3.5:27b
spring.ai.ollama.embedding.model=qwen3.5:27b

# PgVector Configuration
spring.ai.vectorstore.pgvector.enabled=true
spring.ai.vectorstore.pgvector.table-name=vector_embeddings
spring.ai.vectorstore.pgvector.distance-type=cosine
```

## API Endpoints

### Add Document
- **Endpoint**: `POST /api/vector/add`
- **Parameters**:
  - `content`: The text content to embed and store
  - `metadata` (optional): Additional metadata as key-value pairs
- **Response**: Success message

### Search Similar Documents
- **Endpoint**: `GET /api/vector/search`
- **Parameters**:
  - `query`: The search query
  - `topK` (optional): Number of results to return (default: 5)
- **Response**: List of similar documents with their content and metadata

### Clear All Documents
- **Endpoint**: `DELETE /api/vector/clear`
- **Response**: Success message

## How to Run

1. **Build the project**:
   ```bash
   mvn clean install
   ```

2. **Run the application**:
   ```bash
   mvn spring-boot:run
   ```

3. **Test the API**:
   - Add a document: `POST http://localhost:8080/api/vector/add?content=Hello world`
   - Search for similar documents: `GET http://localhost:8080/api/vector/search?query=Hi there`

## Dependencies

- Spring Boot 4.0.5
- Spring AI 1.0.0-M1
- PostgreSQL Driver
- PgVector Store
- Ollama Client

## Notes

- Ensure that PostgreSQL is running with the PgVector extension enabled
- Ensure that Ollama is running and the qwen3.5:27b model is available
- The application will automatically create the necessary table structure in the database
- Vector embeddings are generated using the qwen3.5:27b model via Ollama