CLUDO

Inspiration

Every year, disasters claim millions of lives while authorities react too late. Citizen coordinates pinpoint crises first - we built CLUDO to turn GPS locations into instant life-saving predictions powered by Gemini 3 + satellite data.

What it does

Enter GPS coords → Fetches satellite data → Gemini 3 analysis → "CRITICAL 9.2/10 wildfire risk" + live weather fusion
→ Community upvotes → Authority work orders auto-generated → Live heatmap + crisis dashboard

How we built it

Frontend: Streamlit Cloud deployment → st.map() + coord picker Backend: FastAPI dual APIs → Gemini 3 Pro + OpenWeatherMap Data: Supabase (realtime upvotes + crisis tracking) Flow: Coords → Satellite proxy → Gemini 3 → Weather API → Action plan Deploy: Streamlit Cloud (live: cludo.streamlit.app)

Challenges we ran into

Coordinate-to-satellite mapping: Proxy API for Earth Engine tiles (no auth needed) Gemini 3 prompt optimization: 8s → 2s response with structured prompts Streamlit realtime: Supabase WebSocket conflicts (fixed with polling) Weather data fusion: Dynamic risk recalculation for wind/temp spikes

Accomplishments that we're proud of

Working GPS coordinate → crisis prediction pipeline Dual API architecture cleanly separates concerns Gemini 3 accurately predicts multiple disaster types Community voting system creates viral engagement

What we learned

Gemini 3 excels at coordinate-based reasoning Streamlit + Supabase = perfect hackathon stack Dual APIs future-proofs for municipal integration Satellite proxies unlock Earth Engine without auth hassle