Create a Custom Mobile Webinar App for iOS and Android

You need a mobile webinar app that rivals Zoom and Webex for niche scenarios — online courses, corporate training, webinars with up to 500 attendees. But ready-made solutions lack control over interface, monetization, and data collection. We specialize in mobile webinar app development for iOS and Android using WebRTC and Livekit, building custom apps that give you full control while handling unstable networks, screen sharing, chat with reactions, flexible roles, recording, and playback — all with under 500 ms latency. Over 7 years, we have delivered 20+ such projects, mastering WebRTC, Livekit, Agora, and custom SFU builds. A common mistake is trying to adapt Jitsi Meet to a custom UX, leading to version conflicts and slow development. We avoid this by selecting the right SFU for each task.

Which Video SDK Should You Choose for Your Webinar App?

Building a video conferencing app from scratch on WebRTC is possible but takes 3-6 months just for the media engine. For most products, a specialized SDK that covers SFU, codecs, adaptive bitrate, and network congestion is the right choice.

Livekit SDK — open source, self-hosted SFU, MIT license. iOS/Android/Flutter SDK are available. Standard for new projects where data control matters. Room, LocalParticipant, RemoteParticipant — simple object model. Livekit is 2-3 times cheaper than Agora for 50 participants on 20 hours of webinars per month. Our clients typically save $5,000-$15,000 per month on streaming costs by choosing Livekit over cloud providers.

Agora SDK — powerful but cloud-based with per-second billing. Good for global CDN and minimal DevOps. However, with 100 participants, cost can be 2-3 times higher than self-hosted Livekit.

Daily.co, 100ms, Whereby — cloud alternatives with different pricing. Twilio Video — mature API, but Twilio froze development of Video in 2023.

For self-hosted: Livekit, Janus, and MediaSoup are the three most actively maintained SFUs. Source: WebRTC and Livekit official docs.

Key Features and Implementation Details

Adaptive Video for Mobile Networks

The main challenge for mobile webinars is switching between Wi-Fi and LTE mid-stream. SFU with simulcast solves this: the speaker sends three qualities (720p/360p/180p), and the server selects the optimal one for each listener based on bandwidth.

Livekit: simulcast enabled via VideoPublishOptions(simulcast: true). iOS: LKLocalVideoTrack with CameraSource. Android: LocalVideoTrack + Camera2Source. Automatic resolution reduction on packet loss is built into the SFU.

Dynacast (Livekit) — automatically pauses video tracks of speakers no listener is watching. Saves server bandwidth with >50 participants.

Audio: Opus codec, 48 kHz, AEC + NS + AGC built into WebRTC. Speakerphone vs earpiece: AVAudioSession.overrideOutputAudioPort(.speaker) / AudioManager.setSpeakerphoneOn(true). Automatically switch to speakerphone when >2 participants, earpiece for 1-on-1.

Screen Sharing

iOS (ReplayKit). RPScreenRecorder.shared().startCapture(handler:) — works from iOS 11, accessible from Control Center on iOS 12+. Capture returns CMSampleBuffer — feed into WebRTC RTCVideoCapturer via a custom RTCVideoCapturerDelegate. Limitation: in RPBroadcastSampleHandler (Extension), cannot directly call main app code — use App Group + CFMessagePort or Darwin notifications for state transfer. iOS 15+ RPSystemBroadcastPickerView — native button for starting broadcast without code.

Android (MediaProjection API). MediaProjectionManager.createScreenCaptureIntent() → user confirmation → VirtualDisplay → frame capture via ImageReader → VideoEncoder → WebRTC VideoSource. Requires Foreground Service with type mediaProjection (Android 10+). In manifest: <service android:foregroundServiceType="mediaProjection"/>.

Flutter: flutter_screen_capture (native bindings to both APIs) or a native plugin with platform channels.

Chat, Reactions, Q&A

Real-time chat: WebSocket or Livekit DataChannel (LocalParticipant.publishData). DataChannel uses the same RTC connection — no separate WebSocket server needed.

Reactions (👏🔥❤️): instant animation for all participants. Send via DataChannel, render with particle animation over video grid. iOS: UIViewPropertyAnimator + CAEmitterLayer. Android/Compose: AnimatedVisibility + LaunchedEffect.

Q&A: question queue with upvotes. Storage on server (Livekit RPC or separate REST). Organizer sees a prioritized list.

Roles and Permissions

Organizer, co-organizer, speaker, listener. Difference in permissions: listener does not publish video/audio (saves bandwidth), can only raise hand → organizer grants floor.

Livekit: ParticipantPermission — canPublish, canPublishData, canSubscribe. Set server-side when creating token or changed via RoomServiceClient.updateParticipant.

"Raise hand": DataChannel message {type: "raise_hand", participantId: "..."} → organizer sees an indicator beside the participant.

Recording Webinars

Livekit Egress: cloud recording to S3/R2. EgressServiceClient.startRoomCompositeEgress() — records composite view (all participants in grid) or track-by-track (individual tracks for editing). Format: MP4 with H.264 + AAC.

Access to recording: presigned S3 URL with 24h TTL. In-app player (AVPlayer / ExoPlayer) with positioning and playback speed.

What’s Included in Our Work

• Technical specification and architecture documentation
• SFU setup (Livekit/Janus) and DevOps (K8s/Docker)
• iOS (Swift/SwiftUI) and Android (Kotlin/Compose) app development
• Push notification integration (APNs/FCM), deep linking (Universal Links/App Links)
• User Auth setup (OAuth 2.0, Firebase Auth)
• User onboarding and training guides
• App Store and Google Play support: provisioning profiles, code signing, compliance with guidelines (Section 4.2/5.1)
• Load testing with 50+ participants
• Source code, CI/CD pipelines, and 30 days of post-launch support

Comparison of SFUs for Webinars

Timeline and Process

Our typical process:

1. Requirement analysis and architecture design
2. SFU selection and server setup
3. Core development (video/audio room, roles, screen share, chat)
4. Integration of recording, analytics, and push notifications
5. Load testing and optimization
6. App store submission and launch

Estimated timelines:

• MVP (video room, chat, roles, screen share): 6-8 weeks
• Full-featured (plus registration, analytics, email campaigns): 10-12 weeks

Note: Costs are determined after a detailed analysis of your requirements. We provide a fixed price after the audit phase. Typical MVP cost ranges from $10,000 to $50,000 depending on complexity.

Common Mistakes to Avoid

1. Using Jitsi Meet as a base: It's great for open-source video conferencing but hard to customize for a branded, monetizable app. Version conflicts and slow feature additions are common.
2. Ignoring network variability: Without simulcast and adaptive bitrate, users on mobile networks will experience freezing and disconnections.
3. Overlooking App Store guidelines: iOS apps with in-app purchase or screen recording must comply with strict rules. We handle compliance during development.

Contact us to discuss your webinar app project — we can help you choose the optimal stack and stay within budget. With 7+ years on the market and a proven track record, we guarantee a reliable and scalable solution.