A Flutter developer can build a polished screen while struggling with the native feature your product depends on. Before hiring for background work, device access, or platform SDK integration, inspect the boundary between Dart and the operating system.
The useful question is how the candidate handles lifecycle, errors, and recovery, not how much native code appears in the repository.
Ask the candidate to explain the bridge contract
Flutter's platform-channel documentation provides the communication model. A real integration also needs input types, result states, cancellation meaning, and ownership of native resources.
| Evaluation question | Evidence to request |
|---|---|
| What crosses the bridge? | Explicit input and output contract |
| What can fail? | Native error mapping and user recovery |
| What survives lifecycle changes? | Durable or queryable operation state |
| What is platform-specific? | Reasoned boundary between shared and native logic |
An unexplained Boolean success response can hide several different stages of work.
Discuss an operating-system constraint
For a fictional upload feature, ask what happens when the app backgrounds or the process ends. A method channel does not itself grant unlimited background execution.
The candidate should identify the appropriate platform facility, its limits, and what needs physical-device testing. They do not need to memorise every API signature to explain the architecture responsibly.
The iOS background article gives a more detailed example of this distinction.
Test both sides of the boundary
Ask how native logic and Dart-side behaviour are tested independently, then how the full feature is verified on target devices. Include permission changes, interrupted connectivity, repeated requests, and cancelled operations.
Look for evidence that obsolete callbacks cannot overwrite a newer screen or signed-in account. Resource cleanup matters when repeated use gradually changes memory or device behaviour.
Avoid a rewrite decision based on style alone
A native integration may need repair without moving the whole application into Swift or Kotlin. Ask which specific responsibility belongs natively and why.
My mobile engineering case study describes work across Flutter and native layers. Its relevance is the boundary work, not a claim that every plugin or device scenario has already been tested.
Use a bounded paid proof of concept when the project's critical native capability remains uncertain. Define supported devices and lifecycle tests before estimating the broader feature.
Native-experience questions
Does a plugin dependency prove native expertise?
Ask how the developer would diagnose and maintain it when behaviour breaks.
Should every feature use platform channels?
Use a native boundary when the requirement and supported APIs justify it.
What should a hiring brief include?
List devices, permissions, lifecycle conditions, and the required outcome in a technical brief.