PaperPath

PaperPath is a full-stack research paper learning platform that helps users discover, save, read, and interact with academic papers through an AI-powered tutor experience. The project combines a React frontend, an Express backend, MongoDB storage, and local AI inference for paper understanding and question answering.

1. Project Overview

PaperPath is designed for students, researchers, and curious learners who want to:

browse and explore research papers,
save papers to a personal library,
track what they have read,
upload datasets in PDF or CSV format,
ask questions about papers through an AI tutor.

The application is split into two main parts:

Frontend: React + Vite + Tailwind CSS
Backend: Node.js + Express + MongoDB

2. Features

User features

User registration and login
JWT-based authentication
Personal saved papers library
Reading history tracking
Dashboard for browsing papers
Reader experience for individual papers

AI and content features

PDF ingestion support
CSV ingestion support
Automatic paper metadata storage
AI-powered paper Q&A using local embeddings and generation
Search-like semantic retrieval over stored paper chunks

Admin/Developer features

REST API structure for frontend integration
Upload endpoint for datasets
MongoDB-backed persistence
Easy deployment for frontend and backend separately

3. Tech Stack

Frontend

React 18
Vite
Tailwind CSS
Lucide React

Backend

Node.js
Express.js
MongoDB with Mongoose
JWT for authentication
Multer for file uploads
dotenv for environment configuration

AI Layer

Local Ollama server
Embedding model: nomic-embed-text
Generation model: qwen2.5:0.5b

4. Project Structure

PaperPath/
├── Paper_Path_Backend/
│   ├── server.js
│   ├── package.json
│   ├── src/
│   │   ├── config/
│   │   ├── controllers/
│   │   ├── middleware/
│   │   ├── models/
│   │   ├── routes/
│   │   └── services/
│   └── uploads/
├── Paper_Path_Frontend/
│   ├── index.html
│   ├── package.json
│   ├── public/
│   └── src/
│       ├── components/
│       ├── services/
│       ├── App.jsx
│       ├── main.jsx
│       └── index.css
├── .env.example
├── RagAractictura.md
└── Readme.md

5. Prerequisites

Before running the project locally, make sure you have:

Node.js 16 or newer
npm or pnpm
MongoDB Atlas account or a local MongoDB instance
Ollama installed and running locally
A browser to view the frontend

Install Ollama

If you want the AI features to work, install Ollama and pull the required models:

ollama pull nomic-embed-text
ollama pull qwen2.5:0.5b

Then make sure Ollama is running on:

http://localhost:11434

6. Installation Guide

Step 1: Clone the repository

git clone <your-repo-url>
cd PaperPath

Step 2: Install backend dependencies

cd Paper_Path_Backend
npm install

Step 3: Install frontend dependencies

cd ../Paper_Path_Frontend
npm install

7. Environment Variables

Create a .env file inside the backend folder based on the example file:

cp ../.env.example .env

Example values:

PORT=3001
MONGO_DB_URL=your_mongodb_connection_string
JWT_SECRET=your_secret_key
GEMINI_API_KEY=your_gemini_key_if_needed
VITE_API_BASE_URL=http://localhost:3001/api

Notes

MONGO_DB_URL is required for database connectivity.
JWT_SECRET is required for authentication.
VITE_API_BASE_URL is used by the frontend to call the backend.
The current AI flow uses local Ollama rather than Gemini, so local Ollama must be running for AI features to work.

8. Running the Project Locally

Start the backend

cd Paper_Path_Backend
npm run dev

The backend will run on:

http://localhost:3001

Start the frontend

Open a second terminal:

cd Paper_Path_Frontend
npm run dev

The frontend will run on:

http://localhost:3000

9. API Overview

The backend exposes the following API groups:

Authentication

POST /api/auth/register
POST /api/auth/login
GET /api/auth/profile

Papers

GET /api/papers
GET /api/papers/:paperId
POST /api/papers/:paperId/ask
POST /api/papers/upload

User library

GET /api/users/library
POST /api/users/save/:paperId
POST /api/users/history/:paperId

These routes are used by the React frontend through the API service layer.

10. How the App Works

Flow overview

A user registers or logs in.
The frontend fetches papers from the backend.
The user can save papers and view reading history.
The user can open a paper in the reader view.
The AI tutor can answer questions based on paper content and retrieved context.
Uploading a PDF or CSV stores the content in MongoDB and makes it searchable.

11. Frontend Deployment on Netlify

To deploy the frontend on Netlify, use the following settings:

Base directory: Paper_Path_Frontend
Build command: npm run build
Publish directory: dist
Functions directory: leave empty

Environment variable for Netlify

Add this environment variable in Netlify:

VITE_API_BASE_URL=https://your-backend-domain.com/api

Replace your-backend-domain.com with the actual URL of your deployed backend.

12. Backend Deployment Suggestions

The backend can be deployed on services such as:

Render
Railway
Cyclic
Fly.io
VPS or cloud server

Important deployment notes

Make sure MongoDB Atlas is reachable from the deployed backend.
Set your environment variables securely in the hosting platform.
If you deploy the backend, update the frontend environment variable to point to the new backend URL.
If using a cloud deployment, ensure CORS is enabled correctly for your frontend domain.

13. Troubleshooting

Backend not starting

Check whether MongoDB is reachable.
Confirm your .env values are correct.
Make sure dependencies are installed.

Frontend cannot connect to backend

Verify VITE_API_BASE_URL is set correctly.
Confirm the backend is actually running.
Check for CORS issues if frontend and backend are on different domains.

AI features not working

Ensure Ollama is installed and running.
Verify the required models are downloaded.
Check that the backend can reach http://localhost:11434.

14. Future Improvements

Possible enhancements for the project include:

user profile customization,
paper search and filtering improvements,
full-text search and ranking,
better AI summarization,
admin dashboard for managing uploaded content,
improved deployment automation with CI/CD.

15. Credits

This project was built as a full-stack academic paper learning application with a focus on accessibility, search, and AI-assisted learning.

Author and Creator : Karan Kumar Aditya Kushwaha

Name		Name	Last commit message	Last commit date
Latest commit History 38 Commits
Paper_Path_Backend		Paper_Path_Backend
Paper_Path_Frontend		Paper_Path_Frontend
.env.example		.env.example
.gitignore		.gitignore
RagAractictura.md		RagAractictura.md
Readme.md		Readme.md
package-lock.json		package-lock.json

Folders and files

Latest commit

History

Repository files navigation