DocSmart 📑🧠

DocSmart is an innovative Retrieval Augmented Generation (RAG) agent for document analysis built with Streamlit, Deep-Seek-R1:1.5b, and Ollama. It allows you to upload PDF documents, query their content interactively, and receive concise, factual answers based on your document's context.

Features

• PDF Upload & Analysis: Easily upload your research documents in PDF format.
• RAG-Powered Q&A: Retrieve relevant document sections and generate expert-level answers.
• Interactive Chat Interface: Enjoy a chat-based interface with custom styling for an enhanced user experience.
• Customizable & Extendable: Built using popular libraries like LangChain, making it easy to adapt and expand.

Installation

1. Clone the Repository

   git clone https://github.com/yourusername/docsmart.git
   cd docsmart

2. Set Up a Virtual Environment (Optional but Recommended)

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies Make sure you have Python 3.7+ installed. Then, install the required packages with:

   pip install -r requirements.txt

Running the Application

To launch the DocSmart application, run the following command:

streamlit run rag_deep.py

This command will start the Streamlit server and open the application in your default web browser. You can then upload your PDF document and start interacting with the RAG agent.

Usage

1. Upload a PDF Document: Click the upload button to select a PDF file.
2. Ask Questions: Enter your query about the document in the chat input field.
3. Receive Answers: The system will analyze the document and respond with concise, factual answers.

Contributing

Contributions are welcome! If you have suggestions, improvements, or bug fixes, please fork the repository and create a pull request.

License

This project is licensed under the MIT License.

Acknowledgements

• Built with Streamlit
• Powered by LangChain and Ollama
• Thanks to the contributors and the open-source community for their support.