Our client is a Taiwanese digital entertainment group, founded in 1995. Over time, it has evolved into a full eco-technology enterprise. They have a strong footprint in gaming, digital platforms, and beyond.
Key achievements:
- Their flagship game has generated nearly USD 4.23 billion in lifetime revenue. It has over 3 million registered members in Taiwan, 1.11 million daily active users, and more than 400,000 concurrent players. Average revenue per download ranges from USD 279 to USD 406.
- Another major title has reached over 180 million registered users worldwide.
- Across key PC online games, the company has more than 10 million registered members globally.
- Their virtual payment platform app has grown to 1 million users. The e-commerce platform has seen nearly 40% annual growth in buyers.
- A major entertainment festival hosted by the group attracted over 300,000 participants, both online and offline.
The client wanted to build an in-app communication platform to boost real-time engagement and create richer group interactions. They needed:
- Group channels where users could create, manage, and join shared spaces for messages, media, and updates. This required robust tools for managing members and channels efficiently.
- A no-code UIKit to help users quickly customize their communication interfaces without heavy development work.
- Instant voice and video calls that would enable smooth one-on-one or group conversations, all within the app.
Ultimately, the client aimed to strengthen community engagement and make interactions more dynamic and seamless.
Confidential
To implement the group call feature for the mobile demo app, we designed a solution that combines LiveKit’s robust real-time communication capabilities with a custom layer that aligns with Sendbird’s logic and UI.
Key solution highlights:
1. LiveKit Foundation:
- Utilizes LiveKit’s WebRTC technology for real-time voice and video.
- Leverages its SFU architecture to support high performance and scalability in group calls.
- Employs LiveKit’s mobile SDKs to enable core features like audio, video, and screen sharing.
- Uses LiveKit’s data channels as the backbone for group messaging.
2. Sendbird UI & Logic Alignment:
- Since Sendbird’s UIKit does not natively support calls, we developed a fully custom UI.
- This UI closely mirrors Sendbird’s design principles and overall user experience to maintain consistency.
- A dedicated logic layer was built to map LiveKit’s technical primitives to Sendbird’s conceptual models, ensuring seamless integration.
3. Group Channels & Messaging:
- Group channels are implemented over data channels, supporting text, media sharing, and real-time state synchronization (like typing indicators).
4. “No-Code UIKit” Interpretation:
- This refers to a modular, component-based code structure that developers can easily adapt and customize—rather than a drag-and-drop interface for end-users.
5. Performance Optimization:
- Integrated adaptive streaming, smart codec choices (Opus for audio, H.264/AV1 for video), and noise cancellation techniques to ensure smooth, high-quality performance on mobile devices.
The demo app, built on LiveKit and carefully aligned with Sendbird’s UI and logic, showcased strong performance and versatile in-app communication features.
1. Flexible voice and video call support:
- Seamlessly handles both 1:1 and group calls, supporting up to 20 concurrent users as required for the demo.
- Achieves ultra-low latency, under 100ms across global regions.
- Designed for high availability, ensuring 99.99% uptime for real-time communication.
2. Rich in-app interaction capabilities:
- Group channel messaging lets users create, manage, and participate in channels for text, media, and updates.
- Optimized UI delivered 40% faster response times for messaging.
- Predictive text reduced typing time by 33%.
- Real-time typing indicators and read receipts enhance engagement.
3. Robust real-time call features:
- Users can mute/unmute audio, toggle video, and share their screens for collaboration.
- They can invite others to join or remove participants, with fine-grained permission control.
4. Superior audio quality:
- Uses the Opus codec, scaling from 6 kbps to 510 kbps to maintain excellent voice quality even in low-bandwidth scenarios.
- Improved voice recognition accuracy by 85% through HD voice codecs and advanced noise filtering.
- Includes silence suppression and noise cancellation, reducing bandwidth usage and enhancing clarity.
Together, these results demonstrate a powerful, scalable in-app communication platform that delivers a seamless, high-quality user experience.
