Tapp is a customer care app for Digi, a Malaysian operator. Built with Clean Swift architecture, it features OCR scanning for vouchers and Malaysian ID cards for user registration.

Tapp App

Overview

Tapp is a comprehensive customer care application developed for Digi, one of Malaysia’s leading telecommunications operators. Built from the ground up using Clean Swift architecture, the app provides users with seamless account management, top-up functionality, and innovative OCR-powered features for voucher scanning and user registration.

Key Features

📱 Customer Account Management

Complete account overview and management
Service plan details and usage tracking
Bill payment and transaction history
Customer support integration

🎫 OCR Voucher Scanning

Advanced optical character recognition for Tapp vouchers
Real-time voucher validation and processing
Automatic balance top-up functionality
Error handling and manual entry fallback

🆔 Malaysian ID Card Recognition

Sophisticated OCR for Malaysian identity cards
Automatic data extraction for user registration
Field validation and error correction
Privacy-focused local processing

💳 Payment & Top-up System

Multiple payment method integration
Secure transaction processing
Real-time balance updates
Transaction confirmation and receipts

Technical Implementation

Clean Swift Architecture

Implemented robust, testable architecture:

VIP Pattern: View, Interactor, Presenter separation
Use Cases: Business logic encapsulation
Data Store: Centralized data management
Routing: Scene-to-scene navigation management

Computer Vision & OCR

Advanced image processing capabilities:

Core ML Integration: On-device machine learning for text recognition
Vision Framework: iOS native computer vision capabilities
Image Processing: Pre-processing for optimal OCR accuracy
Text Validation: Smart validation of extracted text data

Malaysian ID Card Processing

Specialized OCR for Malaysian documents:

Document Detection: Automatic ID card boundary detection
Text Extraction: Precise extraction of name, IC number, address
Data Validation: Format validation for Malaysian ID standards
Privacy Protection: Local processing without data transmission

Solo Development Achievement

As the sole developer on this project, I was responsible for:

Complete App Development: End-to-end development from concept to App Store
Architecture Design: Clean Swift implementation and pattern decisions
OCR Implementation: Custom computer vision solutions for Malaysian market
UI/UX Design: User interface design and user experience optimization
Quality Assurance: Testing, debugging, and performance optimization

Technical Challenges & Solutions

Challenge 1: Malaysian ID Card Complexity

Problem: Malaysian ID cards have unique layouts and fonts requiring specialized OCR. Solution: Trained custom Core ML models specifically for Malaysian document formats.

Challenge 2: Voucher Variety

Problem: Multiple voucher designs and formats across different denominations. Solution: Built flexible OCR pipeline adaptable to various voucher layouts.

Challenge 3: Network Reliability

Problem: Inconsistent network connectivity in various Malaysian regions. Solution: Implemented robust offline capabilities and intelligent retry mechanisms.

Challenge 4: Performance Optimization

Problem: OCR processing affecting app responsiveness. Solution: Implemented background processing with progressive UI updates.

OCR Technology Stack

Text Recognition Pipeline

Image Capture: Camera integration with guidance overlays
Pre-processing: Image enhancement and normalization
Text Detection: Vision framework text detection algorithms
Character Recognition: Core ML custom models for Malaysian text
Post-processing: Validation and error correction algorithms

Data Extraction Accuracy

Voucher Scanning: 95%+ accuracy for voucher code recognition
ID Card Reading: 90%+ accuracy for Malaysian IC data extraction
Error Handling: Graceful fallback to manual entry when needed

Business Impact

User Experience

Registration Speed: 70% faster user onboarding with ID scanning
Top-up Convenience: Eliminated manual voucher code entry
Error Reduction: 80% reduction in voucher entry errors
Customer Satisfaction: High ratings for innovative features

Operational Efficiency

Support Reduction: Fewer support tickets due to user errors
Process Automation: Automated data validation and processing
Cost Savings: Reduced manual verification processes

Project Timeline & Delivery

Development Period: 8 months (solo development)
Architecture Planning: 2 months
Core Development: 4 months
OCR Integration: 1.5 months
Testing & Optimization: 0.5 months

Technical Specifications

Platform: iOS (Swift 5+)
Architecture: Clean Swift (VIP)
Computer Vision: Core ML, Vision Framework
UI Framework: UIKit with custom components
Data Storage: Core Data, Keychain Services
Networking: URLSession with custom network layer
Security: Certificate pinning, biometric authentication

This project represents a significant achievement in solo iOS development, combining advanced computer vision technology with clean architecture principles to deliver a production-ready telecommunications application for the Malaysian market.