How teams decide between native, cross-platform and web, and what that decision costs later.
The best mobile stack for AI coding agents is the one your agent can build and verify on your hardest feature, whether native, React Native or Flutter.
React Native is a simpler way to write native apps: you write React in JavaScript or TypeScript, and it renders real iOS and Android UI components.
Modernize a legacy mobile app one screen at a time by adding new code behind a small contract with the native app and testing upgrades before you expand.
Pick an enterprise mobile stack by testing your hard security and device requirements, a real release path, and the upkeep and staffing it will need.
Cross-platform mobile development builds iOS and Android apps from one codebase using Capacitor, Flutter, Kotlin Multiplatform, .NET MAUI or React Native.
Choose a mobile stack by scoring native, cross-platform and web-wrapper options on time to release, release cadence, cost per feature and time to fix.
How AI coding agents write, test and ship mobile apps, and which stacks they work best on.
Verify AI-written mobile code by checking user outcomes against the exact build you will ship, on iOS and Android, including saved data and failure cases.
Agentic mobile infrastructure is the cloud services an AI coding agent needs to build, sign, run, test, ship and monitor a mobile app, not to host AI models.
Build a mobile app with an AI coding agent: scope one testable feature, start from a project that builds, and check the installed app on iOS and Android.
An AI-built app passes App Store review the same way any app does: real features, no secrets in the bundle, accurate privacy labels and a working demo account.
An agentic mobile workflow lets an AI coding agent read build and test results, change code and retry, while CI runs the builds and a person approves releases.
How to run and share a build before it reaches users, with or without Apple hardware.
Xcode and the iOS Simulator only run on macOS, but you can still build and test an iOS app from Windows or Linux in three ways, each with its own cost.
Share a mobile preview without TestFlight: install an ad hoc iOS build or Android APK once, send JavaScript changes as updates, or use a hosted simulator.
Share an unreleased mobile app with non-developers through a simulator link, an internal build, TestFlight or Play testing, or a preview update.
Mobile app testing checks an iOS or Android app at different scopes, from unit tests to end-to-end journeys, on simulators, emulators and real devices.
How to choose a cloud simulator, when a local one is better, and how coding agents, CI, and reviewers use them.
Start an Expo Simulators session from a Bitrise workflow, drive your iOS Simulator build with agent-device, and save screenshots and recordings as artifacts.
Start an Expo Simulators session from a CircleCI job, install your app, run checks on a cloud iOS Simulator, save the evidence, and stop the session.
Run an iOS Simulator from GitHub Actions on Linux: build with EAS, start an Expo Simulators session, drive the app, save screenshots, and stop the session.
Send a link that opens your new build on a cloud simulator in the browser. Private links open for your Expo project, and public links open for anyone.
Put the burden of proof on the agent: it brings screenshots, a video, and a link to the pull request, and the engineer decides if the proof is enough.
Compare cloud iOS simulators, Android emulators, and real-device clouds for CI and AI coding agents, and see where Expo Simulators fits.
How mobile CI/CD works, and how binaries get built, signed, submitted and reviewed.
To submit an app, upload a signed release build to App Store Connect and Play Console, complete each store's forms, then pass review before release.
Set up mobile CI/CD in five stages, from signed builds to store submission, and see how Bitrise, Codemagic, EAS and GitHub Actions handle code signing.
Mobile CI/CD automatically checks each code change and turns it into a signed iOS or Android build that testers or the app stores can receive.
Over 40% of App Store review issues fall under guideline 2.1 App Completeness. See the guidelines behind common rejections and how to fix or appeal.
How to change a live app safely, from over-the-air updates to staged rollouts and rollbacks.
A staged rollout releases an app update to a small share of users first, then widens it; the App Store, Google Play and OTA services each do it differently.
An over-the-air update delivers new code to a mobile app over the network: a new JavaScript bundle, not a new binary, within limits set by the app store rules.
Fix a production bug while your store build waits in review: request expedited review, flip a kill switch, or ship an over-the-air JavaScript update.
How to measure what users feel, and how to catch a regression before it ships.
Reduce app crashes by making every crash report readable, fixing the issues that hit the most users first, reproducing each one and checking the fixed release.
Mobile observability means using crash reports, timings, events and release data from real devices to explain how your app behaves, even when nothing crashes.
Five production metrics show whether your app feels fast to real users, with the thresholds Apple and Google use and how to instrument each.
Find slow screens by timing every route from navigation to first render and to interactive at P90, per release, then fix the ones over budget.
Set a P90 budget per metric, compare each release candidate against the last production release, and block the rollout when a budget is exceeded.
App startup time is how long your app takes to go from launch to a usable screen, and the number depends on the launch state and the endpoint you time.