Multi-Platform Development at Scale

Building applications that work seamlessly across iOS, Android, Web, and Flutter requires careful planning and architectural decisions. This post shares lessons learned from shipping Shrike Labs products across all platforms.

Shared Architecture Patterns

The key to successful multi-platform development is establishing shared patterns while respecting platform-specific constraints.

MVVM Pattern Across All Platforms

iOS: @StateObject ViewModels + Combine Publishers
Android: Hilt-injected ViewModels + StateFlow
Web: Pinia stores + Composition API
Flutter: Riverpod providers + freezed models

Each platform uses its idiomatic patterns, but the underlying logic remains identical.

State Management Consistency

We discovered that state flows follow the same patterns regardless of platform:

1. Action: User initiates an action
2. Processing: Service processes the request
3. Update: UI state updates reactively
4. Side Effects: Related changes cascade

Backend Contracts

Establishing a clear API contract is critical:

{
  "bills": [
    {
      "id": "uuid",
      "title": "string",
      "congress": "number",
      "status": "enum",
      "urgency": "enum"
    }
  ]
}

When the backend changes this contract, all platforms must be updated simultaneously. This requires:

• Versioning strategy
• Deprecation periods
• Clear communication

Testing Strategy

Each platform needs:

• Unit tests: Service logic (shared code)
• UI tests: Platform-specific screens
• Integration tests: API communication

Deployment Coordination

Web & Backend

Can deploy independently on fast schedules (multiple times per day)

Mobile Apps

Require app store review (24-48 hours), so plan carefully

Release Timeline

Monday: Backend deploys
Tuesday: Web deploys
Wednesday: iOS submitted (24h review)
Friday: Android submitted after iOS approval

Common Mistakes to Avoid

1. Platform-specific business logic: Keep logic in backend/services
2. Inconsistent error handling: Standardize error messages
3. Poor version management: Always version your APIs
4. Skipping QA on any platform: Each platform needs testing

Measuring Success

Track these metrics per platform:

• Crash rate
• API latency (p50, p95, p99)
• Feature adoption
• User retention

Conclusion

Multi-platform development at scale isn't about maintaining identical codebases—it's about maintaining identical experiences while respecting platform strengths.

By establishing shared architecture patterns, clear backend contracts, and coordinated deployment processes, your team can ship features rapidly while maintaining quality across all platforms.