Crowdsourced Disaster Relief Platform

A full-stack web application designed to facilitate disaster response efforts through crowdsourcing. The platform connects individuals in need with volunteers and donors, providing real-time data on available resources and emergency alerts.

Features

• AI-Based Volunteer Matching: Matches aid requests with suitable volunteers using KNN algorithm.
• Resource Management: Real-time resource inventory by region.
• Donation & Request Handling: Interfaces for submitting aid requests and making donations.
• Emergency Alerts: Live and historical alert system for disaster events.
• Secure Firebase Backend: Cloud-hosted NoSQL database via Firestore.
• Mobile-Friendly: Built with Flutter for cross-platform deployment.

Tech Stack

Frontend

• Flutter (Dart)

Backend

• FastAPI (Python)
• Firebase Firestore
• AI Matching: scikit-learn, NumPy, Geopy, Joblib

Other Tools

• Docker (optional deployment)
• VSCode
• Pytest for testing

Architecture

• Frontend interacts with users and posts data to backend APIs.
• Backend handles logic, including AI-based volunteer matching.
• Database stores all structured data like users, requests, donations, alerts, and resources.

Data Models

• Users: Basic login info.
• Requests: Type, location, and description of help needed.
• Donations: Donor info and donation type/description.
• Alerts: Emergency type, description, severity, and date.
• Resources: Inventory tracking by area.

AI Matching

• Uses one-hot encoding and K-Nearest Neighbors (KNN).
• Inputs: Request type, location, urgency.
• Matches with volunteers based on skills, location, and availability.

API Endpoints

• Root: http://localhost:8001/
• Production Match Endpoint: http://localhost:8001/match/{request_id}
• Debug Match Endpoint: http://localhost:8001/debug-match/{request_id}
• Swagger UI: http://localhost:8001/docs
• ReDoc: http://localhost:8001/redoc

Frontend Pages

• Home: Navigation to all features.
• Request Help: Submit aid requests.
• Donate: Monetary or material contributions.
• Emergency Alerts: View disaster warnings.
• Resource Inventory: Check resource availability.
• Sign Up / Sign In: UI for user authentication.

Testing

• READ code_1/README1.txt for how to get started and run the application

Deployment

• Backend: Run with Uvicorn or via Docker. 127.0.0.1:8001/match/101
• Frontend: Flutter web app deployable via standard web server. 127.0.0.1:PORT
• Docker setup provided for containerized deployment.

Security

• Firebase credentials are secured and excluded via .gitignore.
• Placeholder user auth exists in frontend, backend integration pending.

Files and What They Do

Files and What They Are:

• code_1 contains the entire website application, has front and backends and needed files to run it
• 2_data_collection contains a data collection program to populate our data base. It also has test cases of our features
• 3_basic_function_testing contains testing the AI matching aspect of our website
• 4_documentaion has all previous reports, final report, and the presentation powerpoint
• code_1/lib has all the frontend
• code_1/backend has all the backend

Watch Demo

• I used my MAC to record this so I have quicktime player to watch the recording

Team

• Casey Nguyen
• Kevin Pulikkottil
• Andy Jih

---

2025 Group 2 - CS 3354