Skip to content

readme.md

Latest commit

 

History

History
62 lines (43 loc) · 1.62 KB

readme.md

File metadata and controls

62 lines (43 loc) · 1.62 KB
Raw

Data Analysis Copilot

Data Analysis Copilot is an AI-powered tool that helps you analyze data using natural language queries. It combines the power of LLMs with SQL and Python to provide interactive data analysis capabilities.

Features

  • Natural language to SQL query conversion
  • Automated data analysis code generation
  • Interactive Streamlit interface
  • Support for SQLite databases
  • Visualization capabilities

Installation

  1. Clone the repository:
git clone https://github.com/SSahas/data-analysis-copilot.git
cd data-analysis-copilot
  1. Create a virtual environment and activate it:
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
  1. Install dependencies:
pip install -r requirements.txt

Usage

  1. Start the Streamlit app:
streamlit run src/app.py

  1. Upload your SQLite database file through the interface.

  2. Enter your analysis questions in natural language.

  3. View the generated SQL queries, analysis code, and results.

Project Structure

data-analysis-copilot/
├── src/                    # Source code             
├── core/              # Core functionality
├── services/          # Business logic services
|── utils/             # Utility functions
├── requirements.txt       # Project dependencies
└── README.md             # Project documentation

Improvements

  1. Add an intent classification layer to know if the user input needs code generation to perform the data analysis or only querying the database. Adding this layer reduces the execution time and also reduces the resources used.