The challenge
African cities are increasingly exposed to flooding, worsened by rapid urbanization and climate change. The lack of accessible early warning systems prevents effective resource allocation and population protection. The challenge: build a dual-platform AI-driven system to predict flood-prone zones and issue real-time alerts.
Methodological approach
In collaboration with a team at ESATIC, the project followed a structured approach: design of a dual-platform architecture (web and mobile), integration of AI predictive models with meteorological data, real-time risk visualization on maps, and implementation of an early warning notification system.
Technical architecture
Data sources: rainfall forecasts, topographic data and historical flood zones. Processing layer: AI/ML models for risk prediction and zone mapping. Visualization: interactive maps with risk zones colored by severity. Platforms: web dashboard and mobile application for authorities and the public.
The deliverable
A functional dual-platform system with live maps, early warning push notifications and risk assessment models.
My contribution
System architecture, AI model training, full-stack implementation. This was the first project where code served not to learn, but to solve a concrete African problem.
Recognition and impact
Operational validation of the prototype. The project demonstrated the ability to address real-world problems using artificial intelligence applied to African meteorology.
Technical resources
Technology badges are displayed below.
- Python
- Machine Learning
- Weather APIs
- Web mapping
- React