📘 DocuMind

DocuMind is a Retrieval-Augmented Generation (RAG) powered chatbot with Google Docs integration.
It enables users to sign in with their Google account, fetch and add Google Docs into a knowledge base (KB), and interact with a chatbot that retrieves relevant context from those documents to answer queries.

---

✨ Features

• 🔑 Google Authentication (OAuth 2.0)
  Users can securely log in with their Google account.

• 📄 Google Docs Integration
  View all documents in your account and select which ones to add to the chatbot’s KB.

• 🤖 RAG-powered Chatbot

  • Retrieves relevant document chunks.
  • Generates accurate answers based on the user’s Docs.
  • Falls back to general knowledge if the answer is not found in Docs.

• 🗂 Knowledge Base Management

  • Add or remove Docs from KB.
  • Query multiple documents at once.
  • Skip duplicate documents automatically.

• 📝 Summarization (Bonus)
  Ability to generate summaries of selected documents using LLM.

• 🌐 Frontend UI
  Clean React + Tailwind interface with a live chat panel.

---

🛠️ Tech Stack

Backend

• Python
• FastAPI – API framework
• ChromaDB – Vector database
• LangChain – RAG pipeline
• Hugging Face Inference API – Embedding + LLMs

Frontend

• React (JavaScript)
• TailwindCSS – styling
• Chat panel UI for interactive experience

---

📚 RAG Pipeline Flow

1. User Query → Sent to backend with selected KB Doc IDs.
2. Vector Retrieval → Relevant chunks retrieved from ChromaDB.
3. LLM Response → Hugging Face model generates an answer.
4. Fallback → If KB has no relevant info, response is generated from general knowledge.

---

🚀 Getting Started

Prerequisites

• Python 3.12+
• Node.js & npm
• Google Cloud credentials (OAuth 2.0 Client ID & Secret)
• Hugging Face API key

Backend Setup

cd backend
python -m venv venv
source venv/bin/activate   # On Linux/Mac
venv\Scripts\activate      # On Windows

pip install -r requirements.txt
uvicorn app.main:app --reload

Frontend Setup

cd frontend
npm install
npm start

---

📌 Assignment Requirements Checklist

• ✅ Google OAuth2.0 login
• ✅ Fetch Google Docs
• ✅ Select and add Docs to KB
• ✅ RAG pipeline with embeddings + Hugging Face LLM
• ✅ Explicit fallback when no KB answer
• ✅ Interactive web-based chatbot UI

Bonus Implemented

• ✅ Summarization
• ✅ Multi-document querying

---

📦 Deployment

The project can be deployed on any cloud provider:

• Backend → Render / Railway / GCP / AWS
• Frontend → Vercel / Netlify