
MERN Stack Disaster Relief Application
Every year, during winter, floods and cyclones in Bangladesh, underprivileged people suffer. Many donors want to help but do not know who to help. This project is a MERN Stack web platform that connects donors, receivers, volunteers and admin to ensure timely relief distribution.
During critical disasters, manually screening hundreds of help requests created bottlenecks, significantly delaying immediate relief to high-risk victims.
Integrated an AI-powered text analysis model to evaluate request titles, extract severity indicators, and automatically calculate urgency scores for instant prioritization or auto-approval.
Managing four distinct user roles (Receiver, Donor, Volunteer, Admin) created security risks regarding unauthorized data access, privilege escalation, and data mutation (e.g., users altering other users' requests).
Implemented JWT authentication with route guard middlewares, enforcing strict role-based authorization rules across both client-side routes and backend REST API endpoints.
High user traffic during crisis situations led to repetitive inquiry support tickets and reduced platform usability for stressed users needing urgent help.
Built and embedded an AI-powered assistant to deliver real-time platform navigation, automated FAQs, and step-by-step guidance for emergency processes.
Fetching and synchronizing dynamic updates across multiple unique dashboards created unnecessary payload overhead and slow response times under high-frequency updates.
Structured optimized RESTful API endpoints with pagination, precise payload scoping, and centralized state management to handle dynamic status changes smoothly.