Inspiration
During floods and earthquakes, we saw that data exists everywhere — satellites, hospitals, emergency calls, telecom towers — but decisions are still slow. The real problem is not lack of data, but lack of unified intelligence.
We wanted to build a system that combines all signals into one clear decision layer.
What We Learned
- Real-time systems depend on clean data flow.
- Simple design is powerful during emergencies.
- AI must explain its decisions, not just predict.
- Offline communication is critical in disasters.
How We Built It
We created two systems:
- Policymaker Dashboard A minimal map-based interface built with Python (FastAPI) and WebSockets. Gemini analyzes multi-source crisis data and gives real-time recommendations.
- Public App An offline-capable app that shows safe zones, hospitals, rescue teams, and allows SOS even without internet.
Challenges
- Combining different data sources into one structure.
- Making the app work without internet.
- Keeping the interface minimal but informative.
- Ensuring AI reliability with fallback logic.
Vision
Turn scattered crisis data into fast, explainable decisions — and help save more lives during disasters.
Built With
- css
- gemini
- html5
- mapbox/leaflet
- python-(fastapi)
- react
- websockets
Log in or sign up for Devpost to join the conversation.