Unleashing the power of artificial intelligence to revolutionize disaster response
[01]
Project overview.
When disaster strikes, the humanitarian response that follows depends on one thing: information. Who needs help? Where is the damage worst? Getting those answers quickly can determine how many lives are saved. SKAI is the World Food Programme's AI-powered platform that analyses satellite imagery to automatically map building damage after a disaster, giving humanitarian teams a fast, accurate picture of where to act first.
[02]
Challenge & Approach.
When a 7.8 magnitude earthquake struck Türkiye and Syria in 2023, SKAI was still in early development. Through Amplify, a team of Netlighters helped bring it to life. We developed and refined the machine learning pipeline powering SKAI's damage classification system, built the deployment infrastructure for emergency-grade reliability, and integrated SKAI into WFP's operational workflows. We also built SKAI+, a visual web interface that makes SKAI's underlying ML models accessible to all WFP staff, regardless of technical background.
[03]
The Impact.
Won 'AI for Efficiency' prize at AI Summit Paris
~500,000 buildings analysed after Myanmar earthquake
300,000+ buildings assessed after Sri Lanka cyclone
Generates impact analyses in under 48 hours (vs. weeks traditionally)
Temporary settlement detection pilot planned for March 2026
