NovaGen

Loading...

Projects/Sisu Saviya InitiativeHumanitarian / Non-Profit

Case Study

Sisu Saviya Initiative

A mission-critical humanitarian platform connecting donors with disaster victims in real-time during Sri Lanka's most challenging moments.

The Challenge

Understanding the Problem

During natural disasters, communication infrastructure often fails. The platform needed to function on extremely low-bandwidth networks (2G/Edge), maintain 24/7 availability during peak crisis periods, and coordinate resources between thousands of donors and recipients simultaneously.

Our Approach

We implemented a Progressive Web App (PWA) with aggressive caching strategies, Incremental Static Regeneration for dynamic content, and a lightweight API designed to work even on the slowest connections.
Phase 1
Offline Map Support with Vector Tiles
Phase 2
AI-based Disaster Impact Prediction
Phase 3
Multi-language Voice Interface
Phase 4
Drone Delivery Coordination

The Solution

Optimized ISR (Incremental Static Regeneration) and service-worker based PWA caching. The app loads in under 2 seconds even on 2G networks and works fully offline.

  • 50K+ Families Reached
  • <2s Load Time
  • 100% Crisis Uptime

The Outcome

Successfully coordinated relief efforts reaching 50,000+ affected families. The platform maintained 100% uptime during the 2023 flood crisis when traditional systems failed.

Impact: 50K+ Families Helped

Technical Deep Dive

Engineering Excellence

A comprehensive look at the technical architecture and implementation details that power this solution.

features

Real-time resource mapping with GPS integration, intelligent donor-recipient matching algorithm, automated SMS alerts via Twilio for non-smartphone users.

scalability

Serverless lambda functions architected to handle 100k+ concurrent users. Auto-scaling infrastructure with zero cold-start latency.

System Architecture

PWA Client

Next.js + Service Worker

CloudFlare CDN

Edge Caching

Serverless API

AWS Lambda Functions

SMS Gateway

Twilio Integration

DynamoDB

NoSQL Database

HTTPS: pwacdn
API Requests: cdnlambda
Send Alerts: lambdasms
Store Data: lambdadb

Development Journey

From Concept to Launch

2 weeks

Crisis Analysis

Conducted field interviews with relief workers and disaster victims to map workflow bottlenecks, identified critical 2G network constraints, and established core requirements for offline-first architecture.

4 weeks

Offline-First Architecture

Implemented Progressive Web App with aggressive service worker caching strategies, configured Incremental Static Regeneration for dynamic donor data, and built fallback mechanisms for zero-connectivity scenarios.

6 weeks

Real-time Features

Integrated Mapbox for GPS-based resource mapping, developed intelligent donor-recipient matching algorithm using geospatial indexing, and connected Twilio API for SMS alerts to feature phones.

2 weeks

Load Testing

Simulated 100K concurrent users with k6 load testing framework, optimized serverless lambda cold starts to <100ms, and fine-tuned auto-scaling policies for crisis traffic spikes.

2 weeks

Deployment & Training

Executed phased rollout starting with pilot district, conducted comprehensive volunteer training sessions, and created multi-language documentation for field workers.

Measurable Impact

Key Results

Primary Outcome
50K+ Families Helped

Direct business value delivered.

50K+
Families Reached
Verified relief deliveries
<2s
Load Time
On 2G networks
100%
Crisis Uptime
During 2023 floods

Impact Analysis

Page Load Time

Before

10s

After

1.8s

5.5x Faster

Crisis Uptime

Before

85%

After

100%

Perfect Reliability

Network Tolerance

Before

3G+

After

2G/Edge

Universal Access

Technology Stack

Tools & Frameworks

Next.jsPWATwilioNode.jsServerlessAWS LambdaDynamoDBMapbox

Implementation

Service Worker Caching Strategy

Aggressive caching with network fallback for 2G reliability.

logic.js
1self.addEventListener('fetch', (event) => {
2 event.respondWith(
3 caches.match(event.request).then((cached) => {
4 // Return cached response immediately
5 if (cached) return cached;
6
7 // Fetch from network with timeout
8 return fetch(event.request, {
9 signal: AbortSignal.timeout(5000)
10 })
11 .then((response) => {
12 // Cache successful responses
13 if (response.ok) {
14 const clone = response.clone();
15 caches.open('v1').then((cache) => {
16 cache.put(event.request, clone);
17 });
18 }
19 return response;
20 })
21 .catch(() => {
22 // Return offline fallback
23 return caches.match('/offline.html');
24 });
25 })
26 );
27});

Performance

Performance Audits

99
Performance
100
SEO
100
Accessibility
95
Best Practices
"This platform saved lives. When everything else failed, Sisu Saviya kept working."
Director, Sri Lanka Red Cross
View All Projects