AgriSmart – AI Farming Assistant

• 

  AgriSmart – AI Farming Assistant

Inspiration

Smallholder farmers in Rwanda often struggle with unpredictable weather, pest outbreaks, and lack of timely farming advice. I wanted to build a tool that uses technology to bridge this gap and empower farmers with real-time guidance.

What it does

AgriSmart provides smart planting schedules, localized weather forecasts, pest alerts, and a voice assistant in Kinyarwanda. It aims to make decision-making easier for farmers with or without internet access.

How I built it

I used Python and TensorFlow to train AI models for pest prediction and image recognition. The frontend was built using React and Flutter, and Firebase handles data storage and user authentication. OpenWeatherMap API delivers weather forecasts.

Challenges I ran into

Training the AI model with limited agricultural data was tough. I also had to simplify the interface so that farmers with low literacy could still use it effectively.

Accomplishments that I'm proud of

I successfully integrated local language voice support and developed a working prototype that is both web and mobile accessible.

What's next for AgriSmart

I plan to test AgriSmart with real users in Eastern Rwanda and integrate more crops and diseases into the AI model. Offline access and SMS alerts are also on the roadmap.

Built With

• api
• firebase
• flutter
• javascript/react
• openweathermap
• python
• tensorflow