EMBER — Emergency Management Body of Evidence & Resources

AI-powered situational awareness for NYC emergency managers. Natural language queries across flood zones, evacuation plans, critical infrastructure, live weather/flood feeds, and uploaded documents — with an interactive operational map.

Backend: Ollama Cloud — fully hosted, no local GPU needed.

---

Quickstart (2 steps)

1. Get an Ollama Cloud API key

Create a free account and generate a key at: https://ollama.com/settings/keys

2. Pull the cloud model

ollama signin
ollama pull gpt-oss:120b-cloud

---

Option A — React/Vite → Vercel

Local dev

git clone https://github.com/YOUR_USERNAME/ember-nyc.git
cd ember-nyc
npm install
cp .env.example .env.local
# Edit .env.local — add your VITE_OLLAMA_API_KEY
npm run dev
# → http://localhost:3000

Deploy to Vercel

npm i -g vercel
vercel login
vercel --prod

Or: connect the GitHub repo at vercel.com → it auto-detects Vite.

Add env vars in Vercel dashboard (Settings → Environment Variables):

Key	Value
VITE_OLLAMA_API_KEY	your_key_here
VITE_OLLAMA_HOST	https://ollama.com
VITE_OLLAMA_MODEL	gpt-oss:120b-cloud

---

Option B — Streamlit → Streamlit Cloud

Local dev

cd streamlit/
pip install -r requirements.txt
export OLLAMA_API_KEY=your_key_here
streamlit run app.py
# → http://localhost:8501

Deploy to Streamlit Cloud

1. Push repo to GitHub
2. Go to https://share.streamlit.io → New app
3. Set Main file path: streamlit/app.py
4. Settings → Secrets — add:

OLLAMA_API_KEY = "your_key_here"
OLLAMA_HOST    = "https://ollama.com"
OLLAMA_MODEL   = "gpt-oss:120b-cloud"

5. Deploy — no server management, free tier available.

---

Architecture

ember-nyc/
├── src/                        # React/Vite (→ Vercel)
│   ├── App.jsx                 # Split-panel: map (left) + chat (right), draggable divider
│   ├── components/
│   │   ├── MapPanel.jsx        # Leaflet dark map — 5 togglable marker layers
│   │   └── ChatPanel.jsx       # Streaming token-by-token chat UI
│   ├── data/nyc.js             # NYC KB, map points, live API endpoints, context builder
│   └── hooks/useOllama.js      # Ollama Cloud streaming client
├── streamlit/
│   ├── app.py                  # Full Python version (Folium map + streaming chat)
│   └── requirements.txt
├── .env.example                # Copy to .env.local, add your API key
├── vercel.json
└── README.md

---

Features

Knowledge Base (NYC pre-loaded)

Module	Contents
Flood Zones	FEMA Zone A/AE/VE/X with post-Sandy context
Evacuation Zones	All 6 zones, shelter locations, contraflow routes
Critical Infrastructure	Trauma centers, ConEd substations, subway flood exposure, water/wastewater
Hazard Profiles	Hurricanes, heat, flooding, winter storms, earthquake, CBRN, pandemic
Emergency Contacts	NYC OEM, FDNY, NYPD, NWS OKX, ConEd, National Grid, FEMA Region 2

Live API Feeds (free, no key needed)

Feed	Source
Active weather alerts	NWS api.weather.gov
7-day forecast NYC	NWS OKX gridpoint
Stream gauge heights	USGS WaterServices
Disaster declarations	FEMA OpenData
Recent 311 reports	NYC Open Data

Map Layers

• 🏥 Trauma Centers (Level 1)
• 🏫 Hurricane Evacuation Shelters
• 📡 USGS/NOAA Stream & Tidal Gauges
• 🏛 EOC / Command Posts
• 💧 High Flood Risk Areas (Zone AE/VE)

Document Ingestion

Drag & drop TXT, CSV, JSON, GeoJSON, MD — ingested immediately into LLM context.

---

Extending EMBER

New data source → add to LIVE_ENDPOINTS in src/data/nyc.js + summarizer case.

New map layer → add to MAP_LAYERS in src/data/nyc.js with lat/lng/name/note.

Different model → change VITE_OLLAMA_MODEL. Available cloud models:

gpt-oss:120b-cloud
gpt-oss:70b-cloud

See full list: https://ollama.com/search?c=cloud

---

Environment Variables

React/Vite (.env.local)

Variable	Default	Description
VITE_OLLAMA_API_KEY	—	Required. Ollama Cloud API key
VITE_OLLAMA_HOST	https://ollama.com	Ollama Cloud base URL
VITE_OLLAMA_MODEL	gpt-oss:120b-cloud	Model name

Streamlit (secrets or env)

Same keys without the VITE_ prefix: OLLAMA_API_KEY, OLLAMA_HOST, OLLAMA_MODEL.

---

License

MIT