A Zero-OpEx, Serverless Institutional Knowledge Management System
🚀 Live Demo • Database • 📱 Android App • 📖 Documentation • 🐛 Report Bug • ✨ Request Feature
- 🎯 Overview
- ✨ What's New in v1.5
- 🏗️ Architecture
- ⚡ Quick Start
- 📚 Documentation
- 🎬 Demo & Screenshots
- 🔬 Research Paper
- 🛠️ Technology Stack
- 🌟 Key Features
- 📊 Performance & Scalability
- 📄 License
- 👥 Team
- 🙏 Acknowledgments
NIELIT StudentHub is a revolutionary serverless platform that enables educational institutions to deploy professional-grade student portfolio systems with $0 annual operational costs. Built on innovative "Zero-OpEx Architecture," it combines Google Sheets as a database, Google Apps Script as a backend, and GitHub Pages for hosting—proving that enterprise features don't require enterprise budgets.
Version 1.5 introduces groundbreaking performance optimizations including:
- Client-Side Instant Search with Fuse.js (0ms latency)
- Server-Side Pagination with "Reverse-Range" strategy (O(1) retrieval)
- Background Worker Pattern for trending calculations (8× faster)
- Multi-Level Caching (RAM + Browser + Database)
- Smart Trending Algorithm with time-decay scoring
- Scalability Improvement: From 15,000 to 500,000+ projects capacity
- 🏛️ Educational institutions with limited budgets
- 👨🎓 Student-led projects and campus initiatives
- 🔬 Research labs showcasing work
- 💼 Department-level portfolio systems
- 🌍 Organizations in developing nations
"Robust frugal software engineering principles applied to low-code infrastructure can democratize digital visibility for resource-constrained institutions."
|
Instant Search (Fuse.js)
Reverse-Range Pagination
Background Worker (Cron Job)
Multi-Level Caching
|
Time-Decay Scoring Formula: Features:
Performance Impact:
Profile Likes System
Enhanced Security
Project Categories
Resume Upload
Public Sharing
|
| Metric | v1.4 (Legacy) | v1.5 (Optimized) | Improvement |
|---|---|---|---|
| Search Latency | ~1500ms | ~10ms | 150× Faster |
| Search Bandwidth | ~2MB/query | ~50KB (cached) | 97% Reduction |
| Read Operations | 1 Full Scan (O(N)) | 1 Partial Scan (O(1)) | 99.9% Efficiency |
| Trending Latency | ~2500ms | ~300ms | 8× Faster |
| Comment Aggregation | N queries | 1 query | 20× Faster |
| Max Capacity | ~15,000 projects | ~500,000 projects | 33× Scalability |
| Annual Cost | $0.00 | $0.00 | Maintained |
graph TB
subgraph "🌐 Client Layer"
A[Browser] -->|HTTPS| B[GitHub Pages CDN]
B --> C[HTML/CSS/JS + Fuse.js]
end
subgraph "☁️ Backend Layer"
C -->|REST API| D[Google Apps Script v1.5]
C -->|Image/PDF Upload| E[Cloudinary API]
D -->|Hourly Trigger| F[Background Worker]
end
subgraph "💾 Data Layer"
D -->|CRUD| G[Google Sheets DB]
F -->|Write Cache| G
G --> H[Users Sheet]
G --> I[Projects Sheet]
G --> J[Profiles Sheet]
G --> K[Comments Sheet]
G --> L[Upvotes Sheet]
G --> M[ProfileLikes Sheet]
G --> N[TrendingCache Sheet NEW]
end
subgraph "🔍 Search Layer"
C -->|Fetch Index| D
D -->|Generate & Cache| O[Search Index RAM]
C -->|Client-Side Search| P[Fuse.js Engine]
end
style A fill:#667eea
style C fill:#764ba2
style D fill:#f093fb
style F fill:#fa709a
style G fill:#4facfe
style N fill:#43e97b
style O fill:#fee140
style P fill:#30cfd0
| Component | Technology | Purpose | Cost | v1.5 Enhancement |
|---|---|---|---|---|
| Frontend | HTML5/CSS3/Vanilla JS + Fuse.js | Static SPA with client-side search | $0 | Added Fuse.js integration |
| Backend | Google Apps Script v1.5 | Serverless API with multi-level caching | $0 | Reverse-range pagination, RAM cache |
| Database | Google Sheets (7 sheets) | Relational store (10M cell capacity) | $0 | Added TrendingCache sheet |
| Storage | Cloudinary | Image/PDF CDN with optimization | $0 | Extended to support PDFs |
| Hosting | GitHub Pages | Global CDN delivery | $0 | No change |
| CI/CD | GitHub Actions | Automated deployment | $0 | No change |
| Search | Fuse.js (Client-Side) | Fuzzy search engine | $0 | NEW in v1.5 |
| Cron | Apps Script Time Triggers | Background worker | $0 | NEW in v1.5 |
Total Annual Cost: $0.00 💰
// PHASE 1: Server generates lightweight index (once per 6 hours)
function getSearchIndex() {
const cache = CacheService.getScriptCache();
const cachedIndex = cache.get('search_index');
if (cachedIndex) return cachedIndex; // Return cached version
// Generate minimal index (only essential fields)
const searchData = allProjects.map(p => ({
id: p.id,
title: p.title,
authorName: p.authorName,
tech: p.tech,
category: p.category
}));
cache.put('search_index', JSON.stringify(searchData), 21600); // 6 hours
return searchData;
}
// PHASE 2: Client performs instant search with Fuse.js
const fuse = new Fuse(searchIndex, {
keys: ['title', 'authorName', 'tech', 'category'],
threshold: 0.3
});
const results = fuse.search(query); // Instant, 0ms latency!Benefits:
- Zero server load for search queries
- Instant results (< 10ms typical)
- Scales to millions of queries/day
- Works offline once index loaded
// OLD (v1.4): Load entire database
const allData = sheet.getDataRange().getValues(); // ❌ O(N) - reads 100%
const pageData = allData.slice(startIndex, endIndex);
// NEW (v1.5): Calculate exact range and read only that
const lastRow = sheet.getLastRow();
const endRow = lastRow - ((pageNum - 1) * 20);
const startRow = Math.max(2, endRow - 20 + 1);
const data = sheet.getRange(startRow, 1, numRows, numCols).getValues(); // ✅ O(1) - reads <0.01%Impact:
- Data Read: 100% → <0.01% per request
- Latency: Variable (1-5s) → Constant (0.8s)
- Scalability: 15K → 500K+ projects
// Runs every hour via Time-Driven Trigger
function updateTrendingCache() {
// 1. Calculate scores for ALL projects (once per hour)
const scores = projects.map(p => ({
...p,
trendingScore: calculateScore(p.upvotes, p.comments, p.daysOld)
}));
// 2. Sort and take top 5
const top5 = scores.sort((a,b) => b.trendingScore - a.trendingScore).slice(0, 5);
// 3. Write to cache sheet
writeToCacheSheet(top5);
}
// User-facing API simply reads the pre-calculated cache
function getTrendingProjects() {
return cacheSheet.getDataRange().getValues(); // ~300ms, always fresh
}Impact:
- API Response: 2500ms → 300ms (8× improvement)
- Compute Load: Decoupled from user traffic
- User Experience: Always fast, never blocking
// LEVEL 1: RAM Cache (10 minutes)
const cache = CacheService.getScriptCache();
const cached = cache.get(cacheKey);
if (cached) return cached; // Fastest path
// LEVEL 2: Database Query
const data = performExpensiveQuery();
// LEVEL 3: Cache for next requests
cache.put(cacheKey, JSON.stringify(data), 600); // 10 min TTL
return data;Cache Hierarchy:
- RAM Cache (10 min) - Fastest, volatile
- Browser Cache (localStorage) - User-specific
- TrendingCache Sheet (1 hour) - Pre-calculated results
- Search Index Cache (6 hours) - Compressed metadata
Learn More About Architecture →
✅ Google Account
✅ Cloudinary Account (free tier)
✅ GitHub Account
✅ Basic HTML/JS knowledge
# 1️⃣ Clone Repository
git clone https://github.com/nielitropar/nielitropar.github.io.git
cd nielitropar.github.io
# 2️⃣ Create config.js
cat > config.js << EOF
const CONFIG = {
SHEET_URL: 'YOUR_APPS_SCRIPT_URL',
CLOUDINARY_NAME: 'YOUR_CLOUD_NAME',
CLOUDINARY_PRESET: 'studenthub_preset'
};
EOF
# 3️⃣ Test Locally
python -m http.server 8000
# Visit: http://localhost:8000
# 4️⃣ Deploy to GitHub Pages
git add .
git commit -m "Deploy StudentHub v1.5"
git push origin main- Create Google Sheet with 7 sheets
- Deploy v1.5 Apps Script:
- Copy contents of
google-app-script-v1.5.js - Set up Script Properties with SALT value
- Deploy as Web App (Execute as: Me, Access: Anyone)
⚠️ CRITICAL: Set up Time-Driven Trigger forupdateTrendingCache(Every 1 hour)
- Copy contents of
- Verify 7 sheets created: Users, Projects, Profiles, Comments, Upvotes, ProfileLikes, TrendingCache
🎉 Your site is live at: https://YOUR_USERNAME.github.io
|
|
CICDPipelinesSecuritySetupGuide.md
|
|
docs/PERFORMANCE_OPTIMIZATION_v1.5.md
|
|
QUICK_REFERENCE.md#customization
|
Student Directory |
Project Feed |
Project Detail Page |
User Portfolio |
- ✅ Bottom Navigation - 3-tab mobile interface (Feed, Profiles, Post)
- ✅ Horizontal Trending - Swipeable trending cards
- ✅ Touch-Optimized - Large tap targets & gestures
- ✅ Responsive Grid - Adapts from 320px to 4K
- ✅ Category Chips - Horizontal scroll filters
- ✅ Instant Search - Works offline once index loaded
Title: NIELIT StudentHub: A Zero-OpEx, Serverless Institutional Knowledge Management System
Authors: Lovnish Verma, Sarwan Singh
Institution: NIELIT Ropar, India
Date: January 24, 2026
-
Time-Decay Engagement Heuristic
Score = (Upvotes×2 + Comments×3) / √(DaysOld + 1)Ensures fair trending without old project domination
-
Reverse-Range Pagination Strategy
- Constant-time O(1) retrieval regardless of database size
- Eliminates full-table scans
- Scalable to 500K+ projects
-
Client-Side Search Architecture
- Zero server load for search queries
- 150× faster than server-side implementation
- Scales to unlimited concurrent users
-
Self-Healing Cryptographic Migration
- Automatic password security upgrade on login
- Zero user disruption
- Backward compatibility maintained
-
Background Worker Pattern
- Decouples heavy computations from user requests
- Pre-calculates trending scores hourly
- 8× performance improvement
-
Frugal Information System Framework
- Proof-of-concept for Zero-OpEx architecture
- Serverless micro-service pattern on commodity SaaS
- Enterprise features without enterprise costs
Citations: paper.bib available for LaTeX integration
{
"markup": "HTML5 (Semantic)",
"styling": "CSS3 (Grid, Flexbox, CSS Variables)",
"scripting": "Vanilla JavaScript ES6+",
"search": "Fuse.js 6.6.2 (NEW in v1.5)",
"fonts": "Inter, Space Grotesk (Google Fonts)",
"responsive": "Mobile-first, 320px-4K",
"dependencies": "Minimal (only Fuse.js for search)"
}{
"runtime": "Google Apps Script (V8)",
"api": "RESTful (GET/POST) - 14 endpoints",
"authentication": "SHA-256 + Salt (with auto-migration)",
"database": "Google Sheets (7 normalized sheets)",
"concurrency": "LockService for race conditions",
"caching": "CacheService (RAM, 10min) + localStorage (Browser)",
"cron": "Time-Driven Triggers (hourly background worker)",
"search": "Lightweight JSON index generation (cached 6h)"
}{
"hosting": "GitHub Pages (Global CDN)",
"cicd": "GitHub Actions (Secrets injection)",
"media": "Cloudinary (Image + PDF optimization)",
"version_control": "Git",
"deployment": "Automated via workflow",
"monitoring": "Apps Script Execution Logs"
}- Fuse.js (v6.6.2) - Client-side fuzzy search (NEW in v1.5)
- Cloudinary Upload Widget - Media upload (inline)
- Native Fetch API - AJAX requests
- CSS Grid/Flexbox - Layout (no framework needed)
|
|
| Metric | v1.4 (Legacy) | v1.5 (Optimized) | Improvement |
|---|---|---|---|
| Search Latency | ~1500ms | ~10ms | 150× Faster |
| Search Bandwidth | ~2MB / query | ~50KB (Cached) | 97% Savings |
| Read Operations | 1 (Full Scan) | 1 (Partial Scan) | 99.9% Efficiency |
| Trending Latency | ~2500ms | ~300ms | 8× Faster |
| Comment Aggregation | N queries | 1 query | 20× Faster |
| Max Capacity | ~15,000 Projects | ~500,000 Projects | 33× Scalability |
| API Response Time | 1.2-1.8s | 0.8-1.2s | 25-33% Faster |
| Concurrent Searches | ~50 (server limit) | Unlimited | ∞ Scalability |
| Cost | $0.00 | $0.00 | Maintained |
Test Scenario: 10,000 projects in database, 50 concurrent users searching
v1.4 Performance:
- ❌ Server crashed after 25 concurrent searches
- ❌ Average search time: 1.8 seconds
- ❌ Database reads: 50 full scans (500K rows read)
v1.5 Performance:
- ✅ Zero server failures with unlimited concurrent searches
- ✅ Average search time: 8ms (client-side)
- ✅ Database reads: 1 index generation (cached for 6 hours)
- ✅ Server only served the 50KB index once; browsers handled all 50 searches independently
Capacity Calculation:
- Google Sheets: 10M cells
- Average row: 20 columns
- Theoretical max: 500,000 projects
- At 10 projects/day: 137 years lifespan
- At 50 projects/day: 27 years lifespan
Concurrency:
- Google Apps Script: ~30 simultaneous executions
- Client-side search: Unlimited (each browser independently)
- Peak users (5% of 2,000): 100 concurrent
- Average RPS: Well below throttle threshold
Search Scalability:
- Current index size: ~50KB for 500 projects
- At 5,000 projects: ~500KB (Fast on 4G)
- At 50,000 projects: ~5MB (Acceptable on WiFi)
- At 100,000+ projects: Consider server-side search with inverted index
Ideal For:
- ✅ Campuses with <5,000 students
- ✅ Departments with <2,000 members
- ✅ Non-real-time use cases
- ✅ Budget-constrained institutions
- ✅ High search volume scenarios
Not Suitable For:
- ❌ Real-time chat applications
- ❌ High-frequency trading platforms
- ❌ >10K concurrent write operations
- ❌ Sub-100ms write latency requirements
This project is licensed under the MIT License - see the LICENSE file for details.
✅ Free to use for personal and commercial projects
✅ Modify as needed for your institution
✅ Distribute modified versions
✅ Private use allowed
TL;DR: Use it however you want, just give credit! 🙌
|
Project Supervisor 📧 sarwan@nielit.gov.in Faculty mentor guiding project vision and research methodology |
Lead Developer 📧 princelv84@gmail.com System architecture, v1.5 optimizations, and technical documentation |
National Institute of Electronics & Information Technology (NIELIT), Ropar
For providing the research environment and testing infrastructure.
- Google Apps Script - Serverless backend with caching
- Google Sheets - Cloud database (7 sheets)
- Cloudinary - Media optimization & CDN
- GitHub Pages - Free hosting with global CDN
- Google Fonts - Typography (Inter, Space Grotesk)
- Fuse.js - Client-side fuzzy search library (NEW in v1.5)
- Watson et al. (2013) - Frugal Information Systems
- Heeks & Molla (2009) - Impact Assessment of ICT Projects
- Vare (2021) - Student-Led Sustainability Projects
- Syeda et al. (2025) - Cost-Efficiency of Serverless Approaches
- Y Combinator (2023) - Hacker News Ranking Algorithm
Special thanks to all students and faculty who tested v1.5 and provided feedback on the performance improvements. Your insights were invaluable in achieving the 33× scalability increase.
Choose your path:
|
I want to deploy in 10 minutes |
I want to understand everything |
I need enterprise security |
💬 GitHub Discussions • 🐛 Report Issue • 📧 Email Support
If this project helped you, consider giving it a ⭐️!
It helps others discover this zero-cost solution for their institutions.
Made with ❤️ at NIELIT Ropar • 2026 • Version 1.5
"Empowering the next generation of tech innovators with Zero-OpEx architecture"
