Dynamodb tables
S3 bucket files
SNS alert
Lambda function
S3 bucket
Lambda event
Lambda function test

DisasterTrain360 AI

Inspiration

Disaster preparedness remains a major challenge in many regions because training programs are often managed using fragmented, manual processes. Training providers, government agencies, and citizens frequently lack a unified platform to coordinate disaster readiness activities, resulting in inefficient enrollment, delayed verification, and limited visibility into preparedness levels.

We built DisasterTrain360 AI to modernize disaster preparedness by creating a centralized digital platform where government agencies, training providers, and citizens can collaborate. Our goal was to make disaster training more accessible, transparent, data-driven, and scalable using AWS cloud technologies.

What it does

DisasterTrain360 AI is a full-stack disaster preparedness and training management platform designed for national-scale deployment.

The platform supports three user roles:

NDMA Administrator – Oversees disaster training nationwide, reviews submitted evidence, monitors preparedness insights, and manages reports.
Training Provider – Creates training programs, manages registrations, uploads completion evidence, and tracks program status.
Public User – Discovers available disaster training programs, enrolls, tracks registrations, and downloads completion certificates.

Key features include:

Secure role-based authentication
Disaster training management
Public enrollment system
Secure evidence upload
AI-powered preparedness insights
GIS-based monitoring dashboard
Automated notifications
Analytics and reporting

How we built it

The application uses a modern cloud-native architecture.

Frontend

React 18
Vite
Tailwind CSS

Backend

Spring Boot 3
Java 17

AWS Services

Amazon DynamoDB for scalable NoSQL data storage
Amazon S3 for secure evidence and document storage
AWS Lambda for asynchronous processing
Amazon Bedrock (Titan) for AI-generated preparedness insights
Amazon SNS for notifications

Authentication is implemented using JWT with role-based authorization.

The application follows a REST API architecture where the React frontend communicates with the Spring Boot backend, which integrates with AWS services.

Challenges we ran into

Some of the major challenges included:

Designing a scalable DynamoDB data model for multiple user roles.
Managing secure file uploads using private Amazon S3 buckets and presigned URLs.
Implementing role-based access control across both frontend and backend.
Integrating AWS Lambda with Amazon Bedrock to generate preparedness insights asynchronously.
Coordinating multiple AWS services while maintaining a clean and secure architecture.
Preparing the application for cloud deployment while externalizing all secrets into environment variables.

Accomplishments that we're proud of

Built a production-style disaster preparedness platform instead of a simple prototype.
Successfully integrated multiple AWS services into one application.
Designed a scalable architecture suitable for national-level deployment.
Implemented secure authentication and evidence management.
Created an intuitive user experience for administrators, training providers, and citizens.