My chat app is already feature-complete; what it needs now is pure speed. You’ll dive into the existing Jetpack Compose + Firebase Firestore codebase and squeeze out every millisecond of wasted work, with special emphasis on eliminating jank and UI freezes. Where the pain is most obvious • LazyColumn lists are triggering excessive recompositions, especially in long conversations. • Startup and the first chat open drag because of that same Compose churn. What I expect you to do 1. Profile with Android Studio Profiler, Layout Inspector, and System Trace until the hotspots are proven—not guessed. 2. Refactor state-handling and keying in LazyColumns so items don’t redraw unnecessarily; confirm with Compose metrics that recomposition counts drop. 3. Tune Firestore listeners and pagination so message streams stay real-time without hammering the network or blocking the UI thread. 4. Trim memory use for images and videos; if needed, introduce better caching or down-sampling strategies and back them up with leak checks in LeakCanary or heap dumps. 5. Reduce cold-start and chat-open times to a target we’ll agree on after your initial audit. Acceptance criteria • Smooth 60 fps (or better) scrolling in long chats on a mid-range device. • No skipped frames flagged by Android Studio for key workflows. • Startup and first-chat render meet the new timing target. • Memory footprint stays stable after 10 minutes of continuous use. • A concise report summarising changes, before/after metrics, and any follow-up tips. You’ll be working in a mature codebase built entirely with Kotlin, Jetpack Compose, Firebase Firestore, Coil, and Paging 3. If that stack feels like home and you’ve spent years shaving frames off real-time apps, I’m ready to hand you the repo and profiler traces so we can get started.