Inspiration
In India, farming isn’t just a profession – it’s a lifeline for millions of families. Yet, farmers often get left out of conversations about new technology, especially AI. We were inspired by a simple question: “What if farmers could be active participants in shaping the AI that affects their crops and livelihoods?”
What it does
Green Loop creates a feedback cycle between farmers and AI models:
- Farmers provide local knowledge (like soil behavior, crop cycles, weather observations).
- AI analyzes this data with open agricultural datasets to give insights (yield forecasts, pest alerts, irrigation advice).
- Farmers then validate or refine these insights, improving the system continuously.
How we built it
- Frontend: React + Tailwind for a clean, mobile-first farmer dashboard.
- Backend: Node.js + Express for handling data input and model requests.
- AI/ML: Python (scikit-learn, PyTorch) with Kaggle datasets for crop yield and weather predictions.
- Database: Firebase Firestore for storing farmer feedback and inputs.
- Cloud: Deployed on Vercel/Render for accessibility.
Challenges we ran into
Translating technical AI outputs into simple farmer-friendly advice (not every farmer understands “confidence intervals” – they want to know whether to irrigate tomorrow).
- Ensuring trust – farmers are cautious about new tech, so involving them in the feedback loop was essential.
- Working with multi-lingual interfaces, since India has diverse languages and dialects.
What we learned
- AI can be powerful, but it only becomes meaningful when local human knowledge shapes it
- Building with farmers in mind made us think beyond “cool tech” – it forced us to design for clarity, trust, and usability.
- Collaboration matters: the best solutions emerge when domain knowledge (farmers) meets technical knowledge (AI).
What's next for
Green loop – Farmers in the AI Loop *Expand the model with satellite data and IoT sensors to improve accuracy.
- Build a voice-based interface for farmers who are less comfortable with text apps.
- Partner with local agri-communities, NGOs, and government bodies to pilot test the platform.
Built With
- express.js
- firebase
- firestore
- node.js
- python
- pytorch
- react
- scikit-learn
- tailwindcss
Log in or sign up for Devpost to join the conversation.