Anaphora

A personalized Coptic Orthodox audio platform — and a complete rewrite of the original 2015 codebase.

The challenge

Anaphora had been on the App Store since 2015. The codebase had drifted, the crash rate was hurting growth, and several long-asked-for features (Siri, CarPlay, offline playback) couldn't be cleanly bolted on. The decision was to rebuild rather than refactor.

What I built

• Designed and implemented a new iOS codebase from scratch in Swift, on Clean Architecture + MVVM, with RxSwift / RxCocoa for reactive streams and PromiseKit at the boundaries.
• Shipped advanced audio features: SiriKit Media Intents, CarPlay integration, offline playback, and a tuned concurrency / networking layer to keep streams smooth on weak connections.
• Built UI with programmatic UIKit plus selective SwiftUI for new screens — kept the architecture identical across both.
• Set up the release pipeline on Fastlane + GitHub Actions, including TestFlight → App Store automation and Crashlytics monitoring.
• Owned full release cycles: TestFlight cohorts, App Store submissions, and post-launch production triage.

Key decisions

• Programmatic UIKit + SwiftUI side-by-side, not "rewrite to SwiftUI." New screens go SwiftUI; legacy media views stay UIKit. The architecture pattern is identical, so switching layers later is a refactor, not a rewrite.
• RxSwift over Combine. The codebase already had Rx muscle memory; the value of a migration didn't justify the disruption.
• Offline-first audio. Storage and playback were designed assuming the network is unreliable, which is the lived reality for many of Anaphora's users.

Outcome

• 98%+ crash-free sessions
• +83% downloads growth
• 7,000+ App Store impressions (rolling 30d)
• 9.67% App Store conversion rate
• 1,600+ active users across multiple countries

(Sources: Firebase Crashlytics + App Store Connect, 2024.)