HawkinsRAG

A Python package for building Retrieval-Augmented Generation (RAG) systems with HawkinsDB and multiple data source integrations.

Features

• Multiple data source support through specialized loaders
• Efficient text chunking and embedding
• Seamless integration with HawkinsDB
• Flexible configuration options
• Comprehensive error handling

Installation

pip install hawkins-rag

Quick Start

from hawkins_rag import HawkinsRAG

# Initialize RAG system
rag = HawkinsRAG()

# Load document
result = rag.load_document("document.txt", source_type="text")

# Query content
response = rag.query("What is this document about?")
print(response)

Supported Data Sources

HawkinsRAG supports multiple data sources through specialized loaders:

• Text files (txt, pdf, docx)
• Web content (YouTube, webpages)
• Structured data (JSON, CSV)
• APIs (GitHub, Gmail, Slack)
• Databases (MySQL, PostgreSQL)
• And many more!

Configuration

config = {
    "storage_type": "sqlite",  # or "postgres"
    "db_path": "hawkins_rag.db",
    "chunk_size": 500,
    "loader_config": {
        "youtube": {
            "api_key": "YOUR_YOUTUBE_API_KEY"
        },
        "github": {
            "token": "YOUR_GITHUB_TOKEN"
        }
    }
}

rag = HawkinsRAG(config=config)

License

This project is licensed under the MIT License - see the LICENSE file for details.

Documentation

For detailed documentation, visit HawkinsRAG Documentation.