az-outbreak-radar

hackathon 2026 Project

📘 AZ Outbreak Radar — Team Setup Guide 🧠 Project Overview

We are building an AI-powered early outbreak detection system for Arizona that:

Collects self-reported symptoms Uses ML + rules to estimate individual risk Aggregates data into community-level outbreak signals Displays insights through a Streamlit dashboard 📁 Project Structure az-outbreak-radar/ │ ├── app/ # Streamlit frontend (User interface) │ └── app.py │ ├── ml/ # Machine learning + risk engine │ ├── model.py │ ├── risk_engine.py │ ├── features.py │ ├── analytics/ # Community + trend analysis │ ├── community.py │ ├── trends.py │ ├── hotspot.py │ ├── data/ # Mock/synthetic data │ ├── mock_users.py │ ├── sample_inputs.json │ ├── assets/ │ ├── diagrams/ │ ├── screenshots/ │ ├── requirements.txt └── README.md 🧑‍💻 Team Roles & Responsibilities 🧑‍💻 Person 1 — ML + Risk Engine

Folder: /ml

Responsibilities: Train logistic regression model Build risk scoring system Convert user inputs → feature vectors Key files: model.py → ML training risk_engine.py → combines ML + rules features.py → feature engineering 🧑‍💻 Person 2 — Streamlit App (Frontend)

Folder: /app

Responsibilities: Build user interface Input form for symptoms + travel Display results (risk score + explanation) Connect ML + analytics outputs Key file: app.py → MAIN APPLICATION 🧑‍💻 Person 3 — Analytics + Community Intelligence

Folder: /analytics

Responsibilities: Aggregate multiple users Compute community risk trends Detect hotspots Generate insights for dashboard Key files: community.py → aggregation logic trends.py → time-based analysis hotspot.py → region-based risk detection ⚙️ Setup Instructions (DO THIS FIRST)

Clone repo git clone https://github.com/YOUR_USERNAME/az-outbreak-radar.git cd az-outbreak-radar
Open in VS Code code .
Create virtual environment Mac/Linux: python3 -m venv venv source venv/bin/activate Windows: python -m venv venv venv\Scripts\activate
Install dependencies pip install -r requirements.txt
Run the app streamlit run app/app.py 📦 Data Flow (IMPORTANT) User Input ↓ ML Risk Engine (ml/) ↓ Individual Risk Score ↓ Analytics Layer (analytics/) ↓ Community Dashboard Output 🧠 Shared Data Contract (DO NOT CHANGE WITHOUT TEAM) Input format: { "fever": 0/1, "cough": 0/1, "travel": 0/1, "animal_exposure": 0/1, "mosquito_index": float, "location": str } Output format: { "score": float, "category": "Low | Moderate | High", "drivers": [list], "explanation": str } 🚀 Work Rules (VERY IMPORTANT) ✔ Parallel work is required Everyone codes at the same time No waiting for others ✔ Each folder is independent Do NOT edit other folders unless integration phase ✔ Use fake data first UI and analytics should work before full ML integration 🔥 Development Phases Phase 1 (NOW) Everyone builds their module independently Phase 2 Connect ML → UI Phase 3 Connect analytics → dashboard Phase 4 Polish + demo + slides 🏆 Final Goal

A working system that:

Accepts user symptom reports Outputs individual risk Shows community outbreak trends Explains WHY risk is high/low