PDFWhisperer

The latest version of the application has been deployed on streamlit cloud, and can be accessed through this link.

Description

PDFWhisperer is a chatbot designed to extract and process information from PDF files, and to enable interactive chat sessions with the content of the PDFs. This project includes components for:

• PDF parsing
• Managing chat sessions
• Implementing a processing chain to facilitate user interactions with PDF content

Purpose: This project is a hands-on learning experience for exploring and mastering Generative AI tools and technologies. This project gives a hands-on experience on building Gen AI applications using retrieval augmented generation (RAG).

Run on local

1. Install the dependencies using the following command:

pip install -r requirements.txt

2. Fill the .env file with the required credentials. Sample file:

OPENROUTER_API_KEY=""
COHERE_API_KEY=""
GROQ_API_KEY=""
COOKIES_PASSWORD=""

3. Run the application using the following command:

streamlit run main.py

Tools and Technologies

• PyMuPDF: Used for efficient PDF parsing.
• Langchain: Utilized for implementing the chatbot functionality.
• Streamlit: Powers the frontend interface.
• Groq cloud, Openrouter, Cohere: for accessing LLMs

Future Scope

There are various ways to expand the project, such as:

• giving the llm access to internet using function calling
• converting the llm chain into an agent, allowing it to make its own decisions
• adding the ability to chat with multiple pdfs at once

Thanks for checking out the repository, feel free to connect with me on LinkedIn.