AI-Powered PDF to Data Converter

Convert PDF files into structured data using AI-powered extraction and generate insights from the extracted data.

Overview

This project provides a solution for converting PDF files into structured data formats, such as JSON, using advanced AI techniques. It offers various functionalities including generalized data extraction, schema-based extraction, and generating insights from the extracted data.

Functionalities

1. Generalized Data Extraction: Extract all data from PDF files in JSON format.
2. Schema-Based Extraction: Extract data based on predefined schemas for specific document types.
3. Generate Insights: Upload the extracted data in CSV format and generate insights by plotting graphs based on user queries.

Dependencies

The project utilizes the following libraries and tools:

• Langchain: For AI-powered PDF data extraction.
• Gemini-pro: For converting extracted data into JSON format.
• Pydantic: For schema-based data extraction.
• Streamlit: For building interactive web applications.
• PandasAi: For data Visualization.
• GROQ with Mixtral: For getting query based insights from .csv data.

Screenshots

1. Dashboard

2. Extract Data

3. Extracted Data Output

4. Schema-Based Data Extraction

5. Schema-Based Data Output

6. Generate Insights

7. Generated Insights Output

8. Pie Chart

How to Run

1. Clone the repository:

   git clone https://github.com/maheshsathe07/AI-powered-PDF-to-Data-Converter.git

2. Navigate to the project directory:

   cd src

3. Install the required dependencies:

   pip install -r requirements.txt

4. Run the Streamlit app:

   streamlit run main.py

5. Access the app in your browser at http://localhost:8501.

Contributors

• Mahesh Sathe
• Atharva Shinde
• Prem Shejole
• Bhushan Sangle

License

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