Imarika Chatbot
Photo Credit: Strathmore University
Overview
Farmers across Busia and Machakos counties have historically had to rely on regional weather forecasts too broad to be useful for daily farming decisions. Without access to localized data, planting and harvesting times were largely guesswork leaving yields vulnerable to unpredictable weather patterns.
The Imarika system deploys locally produced IoT sensors to collect hyperlocal weather data, feeding it into an automated chatbot that delivers crop-specific advice tailored to each farmer's conditions. To reach farmers without smartphones, the project works through village champions who relay information to their communities, while text-to-speech features support users who cannot read. Educational videos train farmers in climate-smart agriculture techniques, and all agronomic advice is scientifically validated through a partnership with the Kenya Agricultural and Livestock Research Organization( KALRO). The project also prioritizes working with women farmer groups to ensure gender equity in access.
By translating raw sensor data into actionable daily guidance when to plant, when to harvest, how to respond to shifting conditions, the system is already helping farmers make more confident, data-driven decisions. And because the sensors are locally produced, the model is significantly more affordable to scale than traditional weather infrastructure, offering a replicable blueprint for AI-enabled climate resilience across smallholder farming communities in Africa.