SQL AI Agent

An intelligent multi-agent system for database operations, analytics, and visualization, powered by OpenAI GPT.

Key Features

Multi-Agent Architecture

• Database Agent: Schema management and evolution
• ETL Agent: Data import and transformation
• Analytics Agent: Business intelligence and visualization
• Orchestrator Agent: Pipeline coordination

Core Capabilities

• Natural language to SQL conversion
• Automated schema generation and management
• Intelligent data import with referential integrity
• Interactive business analytics dashboards
• Query validation and execution
• Async support

Prerequisites

• Python 3.8+
• PostgreSQL 12+
• OpenAI API key
• pip (Python package manager)

Setup

1. Install PostgreSQL:

   # Ubuntu/Debian
   sudo apt-get update
   sudo apt-get install postgresql postgresql-contrib
   
   # macOS using Homebrew
   brew install postgresql@14

2. Create a database and user:

   # Connect to PostgreSQL
   psql postgres
   
   # In psql console:
   CREATE DATABASE your_database_name;
   CREATE USER your_username WITH PASSWORD 'your_password';
   GRANT ALL PRIVILEGES ON DATABASE your_database_name TO your_username;

3. Clone and install:

   git clone https://github.com/yourusername/sqlagent
   cd sqlagent
   python -m venv venv
   source venv/bin/activate  # or .\venv\Scripts\activate on Windows
   pip install -e .

4. Configure environment:

   # Create .env file
   echo "OPENAI_API_KEY=your_openai_api_key" >> .env
   echo "DB_CONNECTION_STRING=postgresql://your_username:your_password@localhost:5432/your_database_name" >> .env

Usage Examples

1. Basic Schema and Data Import

from src.core.agent import SQLAgent
from src.utils.config import Config

async def main():
    config = Config()
    agent = SQLAgent(config)
    
    # Get schema
    schema = await agent.get_schema()
    
    # Import data
    await agent.import_csv_data(schema, "data/sales_data.csv")

2. Interactive Analytics Dashboard

# Generate sales analytics dashboard
await agent.generate_dashboard(results, query)

3. Natural Language Queries

request = QueryRequest(
    question="What are the top 5 products by total sales?",
    include_explanation=True
)
result = await agent.process_request(request)

Example Dashboards

The system provides several pre-built analytics views:

1. Sales Trends Analysis
2. Product Performance Metrics
3. Geographic Distribution
4. Payment and Order Status Analysis

Access the interactive dashboard at http://localhost:8050 after running the dashboard demo.

Project Structure

sqlagent/
├── src/
│   ├── agents/           # Specialized AI agents
│   ├── core/             # Core functionality
│   ├── database/         # Database operations
│   ├── visualization/    # Dashboard generation
│   └── utils/            # Utilities
├── examples/             # Usage examples
└── tests/               # Test suite

Running Examples

# Basic usage
python examples/basic_usage.py

# Dashboard demo
python examples/dashboard_demo.py

# Complete workflow
python examples/complete_workflow.py

Contributing

1. Fork the repository
2. Create your feature branch
3. Commit your changes
4. Push to the branch
5. Create a Pull Request

Troubleshooting

Common Issues

1. Dashboard not showing:

   • Ensure you're accessing http://localhost:8050
   • Check if the port is available
   • Verify data was imported successfully

2. Data import errors:

   • Verify CSV format matches schema
   • Check database permissions
   • Ensure referential integrity

3. Environment issues:

   • Verify .env file exists and is properly formatted
   • Check OpenAI API key is valid
   • Confirm database connection string is correct

License

MIT License - see LICENSE file for details