Agentic Geographic Information System for humanitarian intelligence.

Angel unifies UN/HDX humanitarian data, computes country-level risk and funding urgency, and serves it to an interactive globe UI with clear green/yellow/red status.

• data/: raw and cleaned datasets
• scripts/: local data utilities (e.g., CSV cleanup)
• frontend/: Next.js app (UI + API routes), deployable on Vercel
• databricks/: notebook and workflow scaffolding for daily ingestion/transforms
• docs/: architecture and operating notes

1. Run cleanup:
   • python scripts/clean_data.py
2. Start frontend:
   • cd frontend && npm install && npm run dev
3. Open:
   • http://localhost:3000
4. Optional env:
   • copy frontend/.env.example to frontend/.env.local

• Daily ingestion from UN/HDX into Databricks tables.
• Country risk scoring with funding and humanitarian indicators.
• Next.js API routes serve geo + risk data to globe client.
• User asks natural language questions that map to filtered views.