An intelligent research assistant platform powered by Google Gemini, Sentence Transformers, and live arXiv data.

---

Overview

PaperPulse is a full-stack AI research platform that accelerates academic and professional literature workflows. Upload research papers, extract structured intelligence, conduct semantic question answering across multiple documents, discover related studies via real-time vector recommendations and live arXiv queries, and generate publication-ready literature surveys and citations — all within a unified, authenticated interface.

Live Demo: paperpulse.vercel.app  |  Backend API: Hosted on Render

---

Features

Module	Description
🤖 Research Assistant	Context-aware RAG Q&A synthesized across all uploaded documents with citation attribution
💡 Recommendation Engine	4 search modes (Topic, Abstract, PDF Upload, Domain Browse) using Sentence Transformer cosine similarity + live arXiv API
📚 Literature Survey Generator	Automated multi-paper surveys categorized by objectives, methodologies, findings and themes
⚖️ Paper Comparison	Side-by-side AI breakdown of models, architectures, datasets, and performance tradeoffs
🔍 Research Gap Analysis	Automated discovery of unexplored problem spaces and future research directions
📈 Trend Analysis	Keyword frequency shifts, emerging topics, and research trajectory projections
📝 Citation Generator	APA 7, MLA 9, Chicago, Harvard, IEEE, and BibTeX — formatted instantly
📂 My Papers Library	Unified manager for uploaded PDFs and saved arXiv recommendations with live search
📤 Export	Export full sessions, notes, and surveys to Markdown, PDF, and DOCX
🔐 Authentication	Firebase Google Sign-in with guest-mode fallback
🌙 Themes	Animated dark / light theme toggle, persisted across sessions

---

Tech Stack

Frontend                   Backend                   AI & Data
────────────────────        ────────────────────       ──────────────────────────
React 18 + Vite 5          FastAPI (Python 3.11)      Google Gemini 3.5 Flash
Context API (Auth,          Uvicorn ASGI server        Sentence Transformers
  Theme, Toast)             Pydantic v2                  (all-MiniLM-L6-v2)
Firebase Auth               python-dotenv              TF-IDF hashing fallback
jsPDF / DOCX export        In-memory VectorStore       Live arXiv Atom/XML API
FontAwesome Icons           Cosine similarity          Browser DOMParser (client)
LocalStorage persistence    CORS-enabled

---

Getting Started

Prerequisites

• Node.js 18+
• Python 3.10+
• A Google Gemini API key — Get one here
• A Firebase project — Create one here

---

1. Clone the Repository

git clone https://github.com/guruprasanth02/PaperPulse.git
cd PaperPulse

---

2. Configure Environment Variables

Copy the example file and fill in your values:

cp .env.example .env

# Gemini API Key
VITE_GEMINI_API_KEY=your_gemini_api_key
GEMINI_API_KEY=your_gemini_api_key

# Firebase (Web App config)
VITE_FIREBASE_API_KEY=your_firebase_api_key
VITE_FIREBASE_AUTH_DOMAIN=your-project.firebaseapp.com
VITE_FIREBASE_PROJECT_ID=your-project-id
VITE_FIREBASE_STORAGE_BUCKET=your-project.firebasestorage.app
VITE_FIREBASE_MESSAGING_SENDER_ID=your_sender_id
VITE_FIREBASE_APP_ID=your_app_id

# Optional: Override backend URL in production
# VITE_API_URL=https://your-backend.onrender.com

In development, VITE_API_URL defaults to http://localhost:8000. In production builds, it automatically falls back to /api on the same domain.

---

3. Run the Frontend

npm install
npm run dev
# → http://localhost:3000

---

4. Run the Backend

cd backend
pip install -r requirements.txt
uvicorn main:app --reload --port 8000
# → http://localhost:8000/docs  (interactive API docs)

---

Project Structure

PaperPulse/
├── src/
│   ├── components/
│   │   ├── ChatContainer.jsx         # Research assistant with RAG & citation tracing
│   │   ├── RecommendationPage.jsx    # Vector-based recommendation with 4 search modes
│   │   ├── DocBrowser.jsx            # Paper library — upload, save, and search
│   │   ├── LiteratureSurveyPage.jsx  # Automated literature survey generation
│   │   ├── ComparePapersPage.jsx     # Side-by-side paper comparison
│   │   ├── ResearchGapPage.jsx       # Research gap discovery
│   │   ├── TrendAnalysisPage.jsx     # Trend and keyword trajectory analysis
│   │   ├── CitationGeneratorPage.jsx # Multi-style citation formatter
│   │   ├── ExportPanel.jsx           # PDF / Word / Markdown export
│   │   ├── UploadZone.jsx            # Parallel multi-file PDF ingestion
│   │   ├── SummaryPanel.jsx          # AI-driven document summarization
│   │   └── SettingsPage.jsx          # User preferences
│   ├── context/                      # AuthContext, ThemeContext, ToastContext
│   ├── hooks/
│   │   └── useSessionPersistence.js  # localStorage session sync with debounce
│   ├── services/
│   │   ├── gemini.js                 # Gemini API client — model cascade + fallbacks
│   │   └── recommendation.js        # Recommendation API + arXiv client-side fallback
│   └── firebase.js                   # Firebase initialization
│
├── backend/
│   ├── main.py                       # FastAPI app — /research, /suggest_questions, /feedback
│   ├── recommendation/
│   │   ├── recommender.py            # RecommendationEngine with vector indexing
│   │   ├── vector_store.py           # In-memory cosine similarity store
│   │   ├── embeddings.py             # SentenceTransformer + TF-IDF fallback
│   │   ├── arxiv_fetcher.py          # Live arXiv Atom/XML query and parser
│   │   ├── routes.py                 # FastAPI router — /recommend, /save-paper
│   │   └── models.py                 # Pydantic request/response schemas
│   └── requirements.txt
│
├── api/
│   └── index.py                      # Vercel serverless function entry point
│
├── Dockerfile                        # Multi-stage production container
├── docker-compose.yml                # One-command local container setup
├── vercel.json                       # Vercel frontend + serverless routing config
├── vite.config.js                    # Vite build config with manual chunk splitting
└── .env.example                      # Environment variable template

---

Deployment

PaperPulse supports three deployment strategies. See DEPLOYMENT.md for the full step-by-step guide.

Strategy 1 (Recommended): Vercel + Render

Deploy the React frontend on Vercel's global CDN and the FastAPI backend on Render's dedicated container — eliminating serverless cold-start limitations for long AI inference tasks.

User Browser
  ├── Static assets  ──►  Vercel (Edge CDN, global)
  └── /api requests  ──►  Render (FastAPI, 24/7 container)

1. Deploy the backend on Render → copy the service URL.
2. On Vercel, set VITE_API_URL to your Render URL and add all Firebase + Gemini keys.
3. Add your Vercel domain to Firebase Console → Authentication → Authorized Domains.

Strategy 2: All-in-One on Vercel

Serves both the SPA and the Python FastAPI serverless function from a single Vercel project — pre-configured with vercel.json and api/requirements.txt.

Strategy 3: Docker Container (Any VPS or Render)

docker compose up -d --build
# App available at http://localhost:8000

---

API Reference

Method	Endpoint	Description
GET	/health	Health check
POST	/research	RAG research Q&A
POST	/suggest_questions	AI-generated follow-up questions
POST	/recommend	Vector-based paper recommendations
POST	/search-topic	Live arXiv + vector topic search
POST	/save-paper	Toggle save/unsave a paper
GET	/saved-papers	Retrieve saved papers
POST	/feedback	Log user feedback on AI answers

Interactive API docs available at http://localhost:8000/docs when running locally.

---

Contributing

Contributions, issues, and feature requests are welcome.

1. Fork the repository
2. Create a feature branch: git checkout -b feature/your-feature-name
3. Commit your changes: git commit -m 'feat: add your feature'
4. Push to the branch: git push origin feature/your-feature-name
5. Open a Pull Request

---

License

Distributed under the MIT License. See LICENSE for details.

---

_{Built with React, FastAPI, Google Gemini, and arXiv open data.}