EAS Observe

Know the impact of every release

Observe is wired into how mobile apps ship. It ties every performance metric to the exact build and update that caused it, so you find regressions in minutes instead of days, with a one-click handoff to AI.

Get Started

Monitoring that knows your build pipeline

Because it’s built across your runtime, builds, and updates, Observe sees more out of the box.

Measure what users feel

TTI, cold and warm launches, frame drops, and bundle load on real devices across every session, now captured automatically with no custom traces needed.

See every build

Every release appears as a marker on the timeline. Hover for the delta, or click to see every event from that build or OTA update.

Handoff to AI

One tap in the dashboard, or pipe JSON from the CLI into Claude Code, Cursor, or any agent of your choice. Real answers from real sessions.

Find who's having a bad time.

Sort any metric slowest-first and drill into a single session. See the device, OS version, country, and update channel. Follow the full event timeline from launch to interactive to find exactly where time was lost. Use <logEvent> to track the moments that matter to your product like checkout, onboarding, key flows, etc, and they show up in the session timeline right alongside built-in metrics.

Your agent gets the full context.

Observe previews the prompt before it copies, so you see exactly what your agent receives: the metric, the affected sessions, the build, and the regression delta. No copy-pasting numbers from a dashboard. No explaining the situation from scratch.

Know what every release did.

Every build and OTA update lands as a marker on the performance timeline. Hover for the delta. Click to see every session from that version. The EAS Update tab shows download and apply time alongside performance metrics, so a slow update rollout never looks like a slow app. Catch a bad release before your users report it.

See which screen is slowing things down.

Most performance tools give you one number for the whole app. With Expo Router Integration, the Navigation tab breaks time to interactive down by route, so you find the slow screen instead of guessing. Multi-select routes to compare them directly across builds and updates.

Find the session that hurts.

P50, P90, and P99 show the full range across every device and session. Budget hardware, older OS versions, spotty connectivity - all captured, automatically, with no custom instrumentation needed.

Observe: Ship faster apps

Observe is a new service we released to track performance metrics in your mobile apps. You'll get metrics like time to interactive (TTI), time to render (TTR), cold launch, warm launch, and bundle load times. You can also see performance for every screen in your app, to see where your users are waiting the most.

Watch Video

Frequently Asked Questions

Have more questions or need human support?

Talk to the team with our contact page or socials.

Sentry tells you something crashed. Datadog tells you something is slow. Neither can tell you that a specific build or OTA update caused it, because they don't have your release pipeline. Observe does. Every metric is already tied to the build and update that introduced it, so root cause takes minutes instead of days. They're complementary tools - Observe owns the EAS pipeline view; the others own error tracking and general APM.

Yes. Set sampleRate between 0 and 1 in your configure() call — 0.4 means 40% of installations are tracked. The decision is deterministic per installation, not per session, so the same devices stay in-sample across releases. You can also toggle dispatchingEnabled at runtime to exclude any subset of devices based on your own logic. Changing either setting requires a release or an EAS Update. See the configuration docs for the full reference.

Measure how it feels

One command. Real metrics in minutes. Built into every build you ship.