Introduction

A leading tractor booking platform serving farmers across the African continent with 1000s of active users. The client wanted to modernize its native Android and iOS apps for both Farmers and Tractor Owners to address slow development & release cycles, app stability issues (Crash-free rate of 94%), high backend API call volumes, outdated design and architecture (MVP + BFF SDK), and limited observability and automation in CI/CD. 

• Migrated from BFF SDK to a microservices-based REST API, for better scalability and cost efficiency

• Applied clean, modular architecture with modern UI frameworks: Jetpack Compose (Android) and SwiftUI (iOS)

• Upgraded tech stacks for performance, offline support, and maintainability (Kotlin/Swift + Retrofit/Alamofire)

• Introduced automated CI/CD, >85% test coverage, and E2E test automation for faster, more reliable releases

• Enabled real-time observability and profiling

• Modern UI 2.0 Modular design with improved user experience

• 94% → >99% Crash free rate improvement

• CI/CD Automation Reduced manual intervention in release cycles

• Faster Bug Detection Via automation → increased delivery velocity

• Lower API Calls Improved app responsiveness

Technological Framework

Why these Setup?

Jetpack Compose and Kotlin deliver a modern, declarative Android UI with offline support, while automated CI/CD through Fastlane and GitHub Actions eliminates manual release processes reducing release cycles from weeks to days.

Why this Setup?

SwiftUI and Clean Architecture principles ensure the iOS app is maintainable and testable at scale, while Core Data enables reliable offline functionality for farmers in areas with poor connectivity.

Why this Setup?

Replacing the BFF SDK with a standardised REST API reduces backend coupling and API call volume, while E2E test automation ensures that every release is validated across both platforms before deployment.

Takeaway

Zapcom modernised a leading African tractor platform's native Android and iOS apps improving the crash free rate from 94% to >99%, implementing CI/CD automation, and establishing a modern, testable Clean Architecture.