dbpa 0.1.0

Database Personal Assistant - An AI-powered database management system with advanced text-to-SQL capabilities

DBPA (Database Personal Assistant)

A natural language interface for database management, powered by AI.

Features

• 🗣️ Natural Language Queries: Interact with your database using plain English
• 📊 Table Management: Easy interface for managing database tables
• 🎓 Query Training: Learn from and improve query generation
• 📝 Error Analysis: Comprehensive error tracking and analysis
• 🔄 Vector Store: Efficient storage and retrieval of example queries
• 🌐 Multi-Model Support: Works with OpenAI and Groq

Installation

pip install dbpa

Quick Start

1. Set up your environment variables:

export OPENAI_API_KEY="your-api-key"
export POSTGRES_HOST="your-host"
export POSTGRES_DATABASE="your-db"
export POSTGRES_USER="your-user"
export POSTGRES_PASSWORD="your-password"

2. Start the DBPA interface:

dbpa start

Usage

Basic Query

from dbpa import DatabaseAssistant

assistant = DatabaseAssistant()
result = assistant.query("Show me all customers who made purchases last month")
print(result)

Table Management

from dbpa import TableManager

manager = TableManager()
tables = manager.get_available_tables()
manager.select_tables(["customers", "orders"])

Query Training

from dbpa import QueryTrainer

trainer = QueryTrainer()
trainer.train("Find total sales by product category")

Configuration

DBPA can be configured using a config.json file:

{
    "ai_model": "gpt-4",
    "language": "en",
    "max_tokens": 1000,
    "temperature": 0.7
}

Contributing

We welcome contributions! Please see our Contributing Guide for details.

License

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

Acknowledgments

• Built with Streamlit
• Powered by OpenAI and Groq
• Uses LangChain for LLM interactions