ClimaCare: Design and Implementation of a Weather-Based AI Assistant for Health Optimization

Abstract

Ajah, C. U., Ituma, C., Adene, G., Gift-Adene, A. U., & Ali, O. S. (2026). ClimaCare: Design and Implementation of a Weather-Based AI Assistant for Health Optimization. Afrisophia: Journal of African Experiment, Thought and Experience, Vol. 3, No. 1, pp. 1–13. https://doi.org/10.5281/zenodo.18936796 Abstract This paper presents ClimaCare, a cross-platform AI assistant designed to mitigate weather-related health risks by integrating real-time meteorological data with personalized health recommendations. Leveraging OpenWeatherMap APIs and GPT-4-based large language models (LLMs), ClimaCare addresses critical gaps in public awareness, personalization, and accessibility in existing weather-health applications. Developed using Agile methodology, the prototype employs Python for backend logic, Tkinter for frontend interfaces, and MySQL for data management. Empirical evaluation with 50 participants demonstrated a 35% increase in user awareness of weather-health linkages (measured via pre- and post-exposure Likert-scale surveys) and a System Usability Scale (SUS) score exceeding 85, surpassing benchmarks. The system aligns with UN Sustainable Development Goals 3 (Good Health) and 13 (Climate Action), offering a scalable solution for preventive e-health in climate-vulnerable regions like Nigeria. The system has several limitations, including reliance on stable internet connectivity and varying levels of digital literacy among potential users. Keywords: Weather AI assistant, health optimization, personalized recommendations, climate variability, large language models, public health preparedness