Mobile App Development for Online School
An online school on mobile is not just "website in WebView". As soon as we add offline course access, adaptive bitrate video, and student progress synced across devices, project complexity jumps to full-fledged edtech product level. Let's break down what's under the hood.
Video — Most Expensive Part of Architecture
Educational video in online school is not YouTube-embed. It requires:
-
DRM protection against downloads: iOS —
FairPlay Streaming, Android —Widevine. Flutter packagebetter_playersupports both via DRM configs, but licensing server setup (KSM for FairPlay or Widevine License Server) — separate infrastructure task - Adaptive streaming: HLS for iOS, DASH for Android, transcoding via AWS Elemental MediaConvert or Cloudflare Stream
-
Offline playback: encrypted download via
flutter_downloader+ AES-256 keys tied to deviceId — so downloaded content won't work on another device
Common mistake: use video_player package without DRM. App Store won't reject it, but first screen recording script will capture all your content.
Progress, Gamification, and Adaptivity
Lesson progress — not just "watched: true". Need granularity: view percentage (user reached 73%), last position for resume, test results. Store locally in Hive or drift (SQLite), sync with server on network recovery — offline-first pattern with connectivity_plus and event queue.
Gamification (points, badges, streaks) works well in edtech and retains MAU. Implementation: server-side gamification-service with event bus, client subscribes to achievement_unlocked events via WebSocket or Firebase Cloud Firestore snapshots(). Achievement animation — Lottie file via lottie Flutter package.
Adaptive learning — more complex. If school wants to recommend next lesson based on test results, need simple ML model or rule-based logic on backend. On Flutter client — just display recommendations, all logic server-side.
App Architecture
Clean Architecture mandatory at this complexity:
lib/
core/ # DI (GetIt), Network (Dio), Storage (Hive)
features/
courses/ # domain / data / presentation
player/ # domain / data / presentation
profile/ # domain / data / presentation
auth/ # JWT + refresh token rotation
State — BLoC (flutter_bloc 8.x): predictable, testable, scales well at 3+ developer teams. Navigation — go_router with deep linking (open specific lesson from email campaign).
Integrations:
| Function | Tool |
|---|---|
| Video streaming | Cloudflare Stream / AWS MediaConvert |
| DRM | FairPlay + Widevine |
| Push notifications | Firebase Cloud Messaging |
| Progress analytics | Amplitude / Mixpanel |
| Payments | Stripe / YooKassa |
| Live webinars | Agora / Zoom SDK |
| Certificates | PDF generation on server, share via share_plus |
Publication and ASO
Online school apps pass App Store review under guideline 3.1.1 — if selling courses inside app, Apple requires In-App Purchase. Workaround — "digital services consumed outside app" — works but requires careful metadata description. Google Play more lenient but also restricts external payment links for certain categories.
Process and Timeline
Stages: audit content model → design data schema → Figma UI/UX → development → QA → beta-test with real students → publication → support.
MVP (courses, video without DRM, tests, progress): 12–16 weeks. Full platform with DRM, offline, gamification, webinars: 24–32 weeks. Cost calculated individually after analyzing content model and scaling requirements.
What's Often Underestimated
Supporting different roles (student / teacher / admin / curator) requires RBAC at API level — not "role check in if-else on client". Client code always reversible. Each endpoint must check permissions independently of what UI shows.