NoteSense – AI Study Notes & Doubt-Solving Assistant

NoteSense is a dual-stack web application designed to solve two core problems for students: disorganized notes and unanswered doubts. It ingests study notes (typed or handwritten/scanned images), automatically clusters them by topic, and provides a RAG (Retrieval-Augmented Generation) assistant that strictly cites the student's own material.

Features

• MERN Stack Web App:
  • Note uploads (images/text) and a beautiful subject/topic dashboard.
  • Custom spaced-repetition (SM-2) flashcard reviewer.
  • Complete user authentication via JWT-protected routes.
• Python/FastAPI AI Microservice:
  • Deep Learning OCR: Uses Tesseract for fast extraction, and seamlessly falls back to a Transformer-based Deep Learning OCR (Microsoft TrOCR) for hard-to-read handwritten notes.
  • Semantic Search & Clustering: Uses sentence-transformers and FAISS for fast vector search. KMeans clustering automatically groups uploaded notes by semantic topics.
  • RAG Doubt-Solver: An AI assistant that answers questions by searching ONLY through your uploaded notes, providing exact citations and snippets.
  • Generative AI: Auto-generates spaced-repetition flashcards and quizzes per topic to aid active recall.

Architecture

React Client (Vite)  <--->  Node.js/Express API  <--->  Python/FastAPI Microservice
     (Frontend)           (Auth, DB, Orchestration)        (OCR, FAISS, RAG, GenAI)

• Database: MongoDB (Mongoose)
• Vector Store: FAISS IndexFlatIP
• Inter-service Auth: Secure internal API keys

Quick Start

1. Backend (Express API)

cd server
npm install
# Ensure MongoDB is running locally
node index.js

2. AI Microservice (FastAPI)

cd ai-service
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
# Set LLM_API_KEY in .env
uvicorn main:app --reload

3. Frontend (React)

cd client
npm install
npm run dev