FEATURED09 / 10

AttendPro

Cross-platform anti-proxy attendance system — Flutter mobile apps (Android + iOS), web dashboard (HTML/CSS/JS), device fingerprinting, geo-fencing, and automated faculty reports.

FlutterDartHTML/CSS/JS Node.jsMongoDBExpressRender

Live Demo (Web) View Source

Demo Credentials

Admin

admin@college.edu

admin123

Faculty

f1@clg

123456

Student

stu1@clg

123456

Solo Project Android + iOS + Web

// Problem Statement

The Challenge

Traditional attendance systems are plagued by proxy attendance — students mark attendance for absent friends. Faculty spend hours compiling manual reports that are unreliable. Existing digital solutions are either web-only or lack anti-fraud mechanisms.

Proxy attendance — one student marks for multiple absent classmates
No mobile native app — web-only solutions have poor UX on phones
No device verification — no way to confirm attendance is from the student's own device
No location validation — students can mark attendance from home
Manual report compilation — faculty manually count and calculate percentages

// My Approach

Engineering Thinking

Built a three-platform system: Flutter for Android + iOS mobile apps with native device access, and a web dashboard using HTML/CSS/JS for faculty administration. This ensures students use native apps (better fingerprinting, GPS access) while faculty manage everything from a browser.

Anti-proxy works via triple verification: device fingerprinting (hardware ID unique per phone), geo-fencing (student must be within campus radius), and time-window validation (attendance only open during scheduled class hours). All three must pass simultaneously.

Faculty reports are fully automated — attendance percentages, defaulter lists, subject-wise breakdown, and CSV exports generated without any manual calculation.

CHOSEN

Flutter (Android + iOS) + Web Dashboard

Single Dart codebase for both mobile platforms. Native device access for fingerprinting and GPS. Web dashboard for faculty.

REJECTED

Web-only PWA approach

PWAs can't access hardware device IDs reliably. Native apps provide stronger device fingerprinting for anti-proxy.

REJECTED

Separate Android (Kotlin) + iOS (Swift)

Double the development effort for the same result. Flutter's cross-platform approach was faster for a solo developer.

// Tech Stack

Why These Technologies

Flutter + Dart

Single codebase for Android + iOS. Native device access for hardware fingerprinting, GPS, and camera.

HTML/CSS/JS

Lightweight faculty web dashboard. No framework overhead — fast load times for report viewing and management.

Node.js + Express

REST API backend handling 500+ concurrent student connections. Non-blocking I/O for high-concurrency attendance marking.

MongoDB

Flexible schema for attendance records. Easy aggregation for attendance percentage calculations and defaulter queries.

Device Fingerprinting

Hardware-level device identification ensures one student = one device. Prevents multi-account proxy attempts.

// Technical Deep-Dive

Key Systems

Anti-Proxy Triple Check

Device fingerprint — hardware ID unique per phone
Geo-fence — GPS must be within campus boundary
Time-window — attendance only during scheduled hours
Duplicate detection — same device can't mark twice per session

Automated Faculty Reports

Auto-computed attendance percentages per student
Defaulter list generation (below threshold %)
Subject-wise attendance breakdown
CSV export per class/section/subject

Flutter Mobile App

Student view: one-tap attendance marking
Faculty view: class roster with present/absent toggles
Material Design 3 UI with platform-adaptive widgets
Offline-first: queues attendance if network drops

High-Concurrency API

500+ simultaneous student connections at peak
MongoDB aggregation pipeline for real-time analytics
Session-based duplicate prevention
Rate limiting to prevent automated abuse

// Challenges & Failures

What Broke & What I Learned

Challenge 01

GPS Accuracy Inside Buildings

Geo-fencing worked well outdoors, but indoor GPS accuracy dropped to 20–50 meters. Students in adjacent buildings were incorrectly marked as "within range." Had to widen the geo-fence radius and add WiFi-based location hints as a secondary signal.

Lesson: GPS alone isn't reliable indoors. Multi-signal location (GPS + WiFi + cell tower) improves accuracy significantly.

Challenge 02

Flutter Build for Both Platforms Solo

Building for both Android and iOS as a solo developer meant dealing with two different build toolchains — Gradle for Android, Xcode for iOS. iOS signing certificates and provisioning profiles were especially painful to configure without a Mac initially.

Lesson: Flutter's "write once" promise is real for code, but build/deploy is still platform-specific. Budget time for platform toolchain setup.

Challenge 03

500+ Concurrent Connections at Peak

When an entire batch of 500+ students tries to mark attendance in the first 60 seconds of a class, the API gets hammered simultaneously. Had to implement request queuing, database connection pooling, and staggered retry logic to prevent timeouts.

Lesson: Spike traffic is harder than sustained traffic. Design for burst patterns, not just average load.

// Results & Impact

What It Achieved

Outcome

3-Platform Coverage

Android + iOS + Web — students use native mobile apps, faculty manage via web dashboard. Single backend serves all three.

Outcome

Anti-Proxy Triple Verification

Device fingerprint + geo-fence + time-window. Three independent checks must pass simultaneously to mark attendance.

Outcome

Zero Manual Reports

Faculty never calculate attendance manually. Percentages, defaulter lists, and CSV exports are fully automated.

AttendPro

The Challenge

Engineering Thinking

Why These Technologies

Key Systems

What Broke & What I Learned

What It Achieved

Screenshots & Demo