Project Type-less

Write smarter prompts. Save tokens. Save energy. Save the planet.

---

🚀 About the Project

Have you ever thought about the hidden environmental cost of AI?
Every query consumes energy, and scaled across millions of users, this translates into significant CO₂ emissions.

We built Type-less at PennApps to solve this challenge.
Our tool helps people communicate more efficiently with AI by compressing prompts into their essential form — without losing meaning.

With every optimized prompt:

• ⚡ Fewer tokens are processed
• 🌍 Less compute → lower CO₂ footprint
• 📊 Users track their savings in real time

---

✨ Features

• 🔹 Prompt Optimizer – NLP/ML algorithms shorten redundant queries
• 🔹 Token & CO₂ Savings Tracker – daily and monthly impact dashboard
• 🔹 MongoDB Storage – saves usage history for cumulative stats
• 🔹 Netlify Hosting – fast, serverless frontend deployment
• 🔹 Clean UI – human-friendly design for easy adoption

---

🛠️ Tech Stack

• Frontend: React, Tailwind CSS
• Backend: Flask (Python)
• Database: MongoDB
• Hosting: Netlify (frontend), Flask backend on local/remote server
• Other Tools: Axios, REST APIs

---

📂 Project Structure

PennApps/
│── backend/           # Flask API (compression + savings logic)
│── frontend/          # React + Tailwind UI
│── database/          # MongoDB integration
│── utils/             # Helper scripts (CO₂ + token tracking)
│── README.md

---

⚡ Getting Started

1️⃣ Clone the repo

git clone https://github.com/hariomsah01/PennApps.git
cd PennApps

2️⃣ Backend Setup (Flask API)

cd backend
pip install -r requirements.txt
python app.py

3️⃣ Frontend Setup (React + Netlify)

cd frontend
npm install
npm start

👉 To deploy frontend:

• Push your frontend/ folder to a branch
• Connect the repo to Netlify
• Netlify auto-builds & deploys your React app

4️⃣ Environment Variables

Create a .env file in backend/:

MONGO_URI=your_mongodb_connection_string

---

📊 How It Works

1. User enters a long prompt in the React UI
2. Text is sent to the Flask backend
3. NLP/ML logic compresses the prompt (removes redundant words, simplifies)
4. MongoDB logs:
   • Tokens before vs. after
   • Estimated CO₂ savings
   • Cumulative daily/monthly impact
5. Netlify-hosted frontend displays results + history

---

🌍 Why It Matters

• AI queries consume energy-intensive computation
• By saving tokens, we reduce energy load on servers
• Even small savings per prompt → large impact at scale

📌 Studies show:

• Training a single large AI model can emit hundreds of tons of CO₂ (Strubell et al., 2019).
• Even inference (day-to-day queries) adds up across millions of users.

---

🎯 Future Scope

• Chrome extension for live prompt optimization
• Gamified leaderboards (who saves most CO₂)
• Visual dashboards with community impact stats
• Advanced transformer-based compression

---

🤝 Contributing

We welcome contributions:

1. Fork the repo
2. Create a branch (git checkout -b feature-name)
3. Commit changes (git commit -m 'Add feature')
4. Push (git push origin feature-name)
5. Open a PR 🚀

---

🏆 Hackathon

Built at PennApps XXVI (2025)

• Devpost Link
• GitHub Repo

---

👨‍💻 Authors

• Hari Om Sah – GitHub | LinkedIn
• Bhavika Kothari – GitHub | LinkedIn
• Sujal Shah – GitHub | LinkedIn
• Sourish Mudumby Venugopal – GitHub | LinkedIn
• Team Type-less 2025

---

Learn More

• React Documentation
• Create React App Documentation

📜 License

Licensed under the University of Pennsylvania License.