03 / Machine learning · IoT · Hackathon
CropAI
Suggests which crops to plant in an area, based on its soil and weather.
We built this at the Reboot the Earth Hackathon. It takes soil and weather readings for an agricultural zone and uses a Random Forest model to suggest the crops most likely to do well there.
- Context
- Reboot the Earth Hackathon
- Stack
- Python · scikit-learn · Flask · PostgreSQL · Next.js · TypeScript · Tailwind CSS
The problem
Investors, exporters and local agriculture offices decide what to plant in a zone without much data about its soil and weather.
What I built
- 01A Random Forest classifier trained on soil values (N, P, K, pH) and climate data (temperature, humidity, rainfall).
- 02Inputs from IoT soil sensors, plus weather data from OpenWeather.
- 03A Flask and PostgreSQL backend, and a Next.js dashboard with separate views for central admins, zone admins and investors.
- 04A chatbot that answers questions about a recommendation.
The result
A working prototype, built during the hackathon.