🌐 Multi-RAG Portal

📜 Overview

Multi-RAG Portal is a web-based tool designed to help you interact more intuitively with PDF resources. By harnessing Retrieval-Augmented Generation (RAG), this system enables you to effortlessly upload PDF documents and pose natural language queries. It seamlessly integrates a FastAPI-powered backend with a user-friendly frontend built using Streamlit. The result is a fluid user experience where you can find information based on the contents of your uploaded files, all through simple, conversational queries.

✨ Key Advantages

• Document Submission: Easily send in your PDF documents for indexing.
• Intelligent Q&A: Ask questions about your PDFs and receive context-rich, human-like answers.
• User-Friendly UI: Enjoy a clean, intuitive interface that simplifies interaction.
• Interactive Frontend: Benefit from an interactive Streamlit-based frontend for enhanced user experience.

⚙️ Requirements

Before starting, ensure you have the following set up:

• Python 3.10.12++
• pip (Python’s package manager)
• Node.js and npm (if you need to adjust frontend components)
• Streamlit: For the frontend interface

📚 PDF Tooling

For processing PDF files, you’ll need poppler-utils:

• Ubuntu/Debian: sudo apt-get install poppler-utils
• macOS: brew install poppler
• Windows: Obtain Poppler from its official source and include its bin directory in your PATH.

🚀 Setup and Installation

1. Clone the Repository

git clone https://github.com/RohmaButt/MultiModal-RAG.git
cd multi-rag-portal

2. Create a Virtual Environment (optional but recommended)

python -m venv venv
source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

3. Install the Required Python Packages

pip install -r requirements.txt

4. Install Additional Dependencies (if needed)

• For PDF processing:

pip install pdf2image pillow

• On Ubuntu/Debian:

sudo apt-get install poppler-utils

• On macOS:

brew install poppler

• On Windows:

Download and install Poppler manually. Add its bin directory to your PATH.

5. Set Up Environment Variables

The application requires an OpenAI API key for generating embeddings and processing queries. You can set this up using a .env file.

• a. Install python-dotenv

pip install python-dotenv

-b. Create a .env File

In the root directory of your project, create a file named .env and add your OpenAI API key:

OPENAI_API_KEY=your-openai-api-key-here"

Usage 🚀

Starting the FastAPI Backend

Activate Your Virtual Environment (if not already activated):

source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

Start the FastAPI Server:

uvicorn api:app --reload

Backend Access:

Open a web browser and navigate to http://localhost:8000 to access the FastAPI backend (primarily used by the Streamlit frontend).

Using the Streamlit Frontend

Ensure the Backend is Running:

Make sure the FastAPI server is active as described above.

Run the Streamlit App:

In a separate terminal window, navigate to the project directory and run:

streamlit run frontend.py

Interact with the Application:

1. Use the Streamlit interface to upload PDFs and submit queries.
2. View responses directly within the Streamlit app, including any embedded images.

API Endpoints 📡

The FastAPI backend provides the following endpoints:

• GET /: Serves the main HTML page (primarily used for backend testing; the Streamlit frontend interacts directly with the API).
• POST /upload_pdf/: Endpoint for uploading PDF files.
• POST /query/: Endpoint for submitting queries.

Detailed Endpoint Descriptions

1. GET /

• Description: Serves the main HTML page.
• Usage: Access via http://localhost:8000/ in a web browser.
• Response: Returns the index.html file located in the static directory.

2. POST /upload_pdf/

• Description: Uploads a PDF file for processing.
• Parameters:
• file (form data): The PDF file to upload.

3. Response:

• Success (200):

{
  "filename": "uploaded_file.pdf",
  "message": "File uploaded successfully"
}

• Error (400):

{
  "detail": "Only PDF files are allowed"
}

3. POST /query/

• Description: Submits a natural language query related to the uploaded PDFs.
• Parameters:
  • question (JSON body): The user's query.

{
  "question": "What is the main result of the paper?"
}

• Response:
• Success (200):

{
  "response": "The main result of the paper is..."
}

• Error (400):

Copy code
{
  "detail": "Please upload a PDF first"
}

📁 Project Structure

MultiModal-RAG/
├── api.py
├── addvectorstore.py
├── main.py
├── load_files.py
├── imgproc.py
├── ragretriever.py
├── retrieval.py
├── frontend.py
├── requirements.txt
├── .env
├── static/
│ └── index.html
├── uploads/
│ └── (uploaded PDFs and extracted images)
└── README.md