title	PDF AI Interaction Backend
emoji	📄
colorFrom	blue
colorTo	indigo
sdk	docker
pinned	false

📄 PDF AI Assistant

A full-stack application that allows users to upload PDF documents and interact with them using natural language. Built with FastAPI, React, LangChain, and Hugging Face.

🚀 Features

• Deployed & Live: Fully hosted on Hugging Face Spaces (Backend) and GitHub Pages (Frontend).

• PDF Upload & Processing: Automatically extracts text, chunks it, and generates vector embeddings.

• RAG Pipeline: Uses Retrieval-Augmented Generation to answer questions based only on the PDF content.

• LLM Integration: Powered by Meta LLaMA 3.1 8B Instruct (via Hugging Face API) for high-quality responses.

• Vector Search: Efficient similarity search using FAISS (with in-memory fallback).

• Interactive Chat UI:

  • Real-time chat interface.
  • Markdown Support: Renders lists, code blocks, and bold text properly.
  • Auto-scroll: Keeps the latest message in view.
  • Persistence: Chat history is saved locally so you don't lose progress on refresh.

• Responsive Design: Beautiful, modern UI with gradient styling.

🛠️ Tech Stack

Backend

• Framework: FastAPI
• Hosting: Hugging Face Spaces (Docker)
• LLM Orchestration: LangChain
• Model Provider: Hugging Face Inference API
• Embeddings: sentence-transformers/all-MiniLM-L6-v2 (Local)
• Vector Store: FAISS (Facebook AI Similarity Search)

Frontend

• Library: React.js
• Hosting: GitHub Pages
• Styling: CSS3 (Custom gradients & animations)
• HTTP Client: Axios
• Rendering: react-markdown

📋 Prerequisites

• Python 3.8+
• Node.js & npm
• A Hugging Face Account & API Token.
• Access to meta-llama/Meta-Llama-3.1-8B-Instruct (Accept license on HF model page).

⚙️ Installation & Setup

1. Clone the Repository

git clone https://github.com/divyansshu/Automating-PDF-Interaction.git
cd "Automating PDF Interaction"

2. Backend Setup

Navigate to the backend folder and install dependencies:

cd backend
# Create virtual environment (optional but recommended)
python -m venv venv
# Windows:
venv\Scripts\activate
# Mac/Linux:
source venv/bin/activate

# Install requirements
pip install -r requirements.txt

Configure Environment Variables: Create a .env file in backend/app/.env:

HUGGINGFACEHUB_API_TOKEN=your_hf_token_here

Run the Server:

cd app
uvicorn main:app --reload

The backend will start at http://localhost:8000.

3. Frontend Setup

Open a new terminal and navigate to the frontend folder:

cd frontend
npm install
npm start

The app will open at http://localhost:3000.

💡 Usage

1. Open the web app.
2. Click "Upload PDF" and select a document.
3. Wait for the processing (extraction & embedding generation).
4. Once done, you will be redirected to the chat screen.
5. Ask any question about your PDF!
6. Use the "Upload New PDF" button to reset and start over.

🛡️ License

This project is open-source and available under the MIT License.