The problem ResQ solves

RESQ is a disaster management platform that brings together AI-driven disaster classification, real-time monitoring, and team coordination tools. It is designed to solve some of the most critical issues in disaster response:

Early Disaster Detection and Classification – The AI model classifies images into four disaster types (Cyclone, Earthquake, Flood, Wildfire), enabling faster and more accurate recognition of disasters.

Real-Time Monitoring (RTM) – It provides live tracking of disasters with important metrics like rainfall, wind speed, temperature, and water levels, helping authorities keep a close watch on changing conditions.

Disaster Prediction – The platform forecasts how disasters may evolve, supporting proactive decisions like evacuation and resource allocation.

Team Coordination – With features for task assignment, priority management, and communication, RESQ ensures smooth coordination between disaster response teams.

Centralized Information – Acts as a single, reliable hub for disaster-related updates and data, making critical information accessible to all stakeholders in one place.

Challenges we ran into

Since this was my first full-stack project with ML integration, there were quite a few challenges along the way:

ML Model Deployment – Setting up the AI classification model using transfer learning (ResNet50), fine-tuning it, and then serving it via Flask was a big learning curve.

Process Management with PM2 – Had to figure out PM2 (through ecosystem.config.js) to keep the Flask server running continuously in production.

Full-Stack Integration – Making the React + TypeScript frontend talk smoothly with the Python ML backend required building a proper API service layer.

Authentication – Implementing Clerk for authentication and securing routes only for logged-in users.

Interactive Maps – Using Leaflet to add disaster visualization on maps for better geographic representation.

Real-Time Data Handling – Building systems that can display live metrics and alerts for ongoing disasters.

Complex UI Components – Designing prediction timelines, team dashboards, and monitoring interfaces with smooth interactivity.

TypeScript Adoption – Learning and applying TypeScript throughout the frontend for stronger type safety.