Lag and packet loss are typical problems when organizing online broadcasts. We build webinar platforms that solve these at the architecture level: from media server selection to CDN streaming. With over 5 years of experience in video platform development and more than 20 successful projects, we guarantee stable streams for audiences ranging from dozens to tens of thousands of participants. If you've experienced lags during webinars or quality degradation when scaling, our architecture eliminates that. We work out CDN topologies, configure HLS segmentation, and optimize WebRTC for minimal latency.
How we build video architecture
P2P (WebRTC without a server) is only suitable for video calls with 3-4 participants. With more, the client load grows exponentially.
SFU (Selective Forwarding Unit) — the media server forwards streams between participants without transcoding. Each participant sends one stream and receives streams from others. Scales to 50-200 participants with latency under 1 second.
MCU (Multipoint Control Unit) — mixes all streams into one. Client load is minimal, suitable for thousands of viewers, but server load is higher.
CDN streaming (HLS/DASH) — the host streams via RTMP, the media server transcodes to HLS, CDN distributes to viewers. Latency 5-30 seconds, unlimited scale.
For webinar format (1 host + thousands of viewers), we use SFU + CDN streaming for scale-out.
| Technology |
Latency |
Max participants |
Client load |
| P2P |
< 100 ms |
3-4 |
High |
| SFU |
< 1 s |
50-200 |
Low |
| MCU |
< 500 ms |
Thousands |
Minimal |
| CDN |
5-30 s |
Unlimited |
Minimal |
What is SFU and why do you need it?
SFU is a server that receives video streams from participants and forwards them to others without transcoding. This reduces client and server load while maintaining low latency. For webinars, SFU enables up to 200 participants with full interactivity.
Media servers
LiveKit — open-source SFU in Go, actively developed. SDKs for React, Vue, iOS, Android. Supports ingress (RTMP, HLS input). Hosting: self-hosted or LiveKit Cloud. LiveKit is a proven solution.
import { Room, RoomEvent } from 'livekit-client';
const room = new Room();
await room.connect('wss://yourinstance.livekit.cloud', token);
await room.localParticipant.enableCameraAndMicrophone();
Mediasoup — lower-level SFU library for Node.js, requires more development but maximum flexibility.
Agora, Daily.co, Vonage — managed services, pay per minute, fast integration, less control.
Comparison of SFU solutions
| Parameter |
LiveKit |
Mediasoup |
Agora |
| Implementation complexity |
Medium |
High |
Low |
| Cost |
Free (self-hosted) |
Free |
Per-minute |
| Flexibility |
High |
Maximum |
Low |
Webinar lifecycle
Creation → Configuration → Promo page → Participant registration → Email reminders → Stream start → Interactive (chat, Q&A, polls) → Completion → Recording processing → Recording distribution → Analytics
Registration and participant management
Webinar landing page: topic, date/time with timezone, speakers, agenda, registration form. After registration — confirmation email with unique token login link.
Reminders: 24 hours, 1 hour, 15 minutes before. Integration with Google Calendar / Outlook via .ics file.
Host tools
- Screen sharing — via WebRTC
getDisplayMedia()
- Whiteboard — collaborative board (Excalidraw-compatible component)
- Slides — presentation via
iframe or PDF viewer
- Hand raise — question signal
- Spotlight — highlight a participant (bring on stage from viewers)
- Breakout rooms — split into rooms for workshops
Recording and processing
After the webinar, recording is processed:
- Media server outputs raw recording (MP4 or WebM)
- FFmpeg transcodes to optimal format with multiple resolutions
- Uploaded to Cloudflare Stream or Mux
- Distributed to attendees (and non-registrants if needed)
Webinar analytics
- Peak concurrent attendance
- Retention: when participants left (% remaining at each moment)
- Engagement: chat activity, poll responses
- Conversion: registrations → attendees → completed view
What's included in the work
- Architectural solution diagram
- Server and client code
- REST API documentation
- Infrastructure setup (Docker, Nginx, CDN)
- Client team training
- 6-month warranty
Timelines
MVP (landing, registration, webinar room with chat, recording, email distribution): 3-4 months. Full platform with multiple speakers, breakout rooms, analytics, CRM integration: 5-8 months.
Order a consultation for your project — we'll assess tasks and propose an architecture. Contact us to discuss details and get an individual development plan.
Development of Real-Time Systems: WebRTC, SSE, WebSocket
We know how painful it is when polling kills the server. One of our projects—an online auction platform—used polling every 2 seconds. Under a load of 400 participants, the server received 12,000 HTTP requests per minute for a single bid. 90% of responses were empty. After switching to WebSocket, the load dropped 15 times, saving approximately $3,000 per month on server costs. Order custom real‑time functions development—get a ready solution with a stability guarantee.
Implementing real‑time in production is not just a library. We design the architecture for load, scenarios, and budget. Below is a breakdown of key solutions with examples.
Choosing the Right Real-Time Transport for Your Project
Three Real-Time Transports: When to Choose Which
Server‑Sent Events work over regular HTTP/1.1 or HTTP/2. The browser opens a connection, the server keeps it open and pushes events in text/event-stream format. Automatic reconnection is built-in—no need for reconnect logic. Limitation: server → client only. Ideal for notifications, progress of long tasks, live feeds.
WebSocket is a full‑duplex channel after an HTTP Upgrade handshake. Browser and server exchange frames in both directions. Suitable for chats, collaborative editing, games, trading terminals. Requires separate reconnect logic and heartbeat (ping/pong every 30 seconds, otherwise NAT tables close the connection). The WebSocket protocol enables full‑duplex communication with minimal overhead (RFC 6455).
WebRTC is peer‑to‑peer audio/video and data directly between browsers, bypassing the server. A server is needed only for signaling (STUN/TURN for NAT traversal). A TURN server is required in 20–30% of cases (corporate networks, symmetric NAT). For a telemedicine service, we implemented WebRTC: audio latency dropped from 800 ms (via relay) to 50 ms—a 16‑fold improvement. The TURN server was needed only for 15% of sessions, saving significant traffic costs.
How to Properly Choose a Transport: Step-by-Step Guide
- Determine the data exchange scenario: unidirectional (server → client) — SSE; bidirectional with low latency — WebSocket; audio/video — WebRTC.
- Evaluate latency requirements. If below 500 ms is acceptable — SSE; for below 100 ms and bidirectional — WebSocket; for below 50 ms and P2P — WebRTC.
- Check the infrastructure budget. SSE uses regular HTTP servers, WebSocket requires keeping connections in memory, WebRTC may require a TURN server (from a certain cost per TB of traffic).
- Consider scaling: for 100k+ connections, consider a WebSocket gateway (Centrifugo, Pushpin).
| Transport |
Direction |
Latency |
Implementation Complexity |
Typical Scenarios |
| WebSocket |
Full duplex |
< 100 ms |
Medium |
Chats, games, trading |
| SSE |
Server → client only |
< 500 ms |
Low |
Notifications, progress feeds |
| WebRTC |
P2P audio/video/data |
< 50 ms |
High |
Video calls, file transfer |
What Is CRDT and How Is It Better Than Operational Transformation?
Collaborative editing is not just "whoever writes last wins". Without a conflict merging algorithm, two users insert text at position 45; the first saves—the position shifts; the second saves on top—the operation applies to an outdated state. Text gets duplicated or lost.
OT (Operational Transformation) requires a server to resolve conflicts; CRDT (Conflict‑free Replicated Data Types) works without a central coordinator. Yjs is the most mature CRDT library for the browser. It integrates with ProseMirror, TipTap, CodeMirror, Monaco Editor. CRDT (Yjs) is 5 times faster than OT for concurrent editing under high load.
Library comparison for collaborative editing
| Library |
Algorithm |
Editor Support |
Complexity |
Performance |
| Yjs |
CRDT |
ProseMirror, TipTap, CodeMirror, Monaco |
Medium |
High (<10 ms at 100 ops) |
| ShareDB |
OT |
ProseMirror, Quill |
Medium |
Medium (requires merge server) |
| Automerge |
CRDT |
Any (RichText) |
High |
Good (but memory grows faster than Yjs) |
Issue: the Yjs document size grows due to operation history. Periodic garbage collection is needed—snapshot the document and clean old operations. Without it, a document worked on for a year may weigh 50 MB.
WebSocket Heartbeat Example (Node.js)
const ws = new WebSocket('wss://example.com');
let pingInterval;
ws.on('open', () => {
pingInterval = setInterval(() => {
ws.ping();
setTimeout(() => {
if (ws.readyState === WebSocket.OPEN) ws.terminate();
}, 5000);
}, 25000);
});
ws.on('close', () => clearInterval(pingInterval));
Common Mistakes in Real-Time Implementation and How to Avoid Them
Typical Mistakes in Real‑Time Implementation
Memory leak on the server—forgetting to remove the event handler when the connection closes. On Node.js, heap grows ~1 MB/hour. EventEmitter warns about 10+ listeners, but it's not always noticed.
Thundering herd on reconnect. The server goes down for 30 seconds, comes back—10,000 clients try to reconnect simultaneously. Exponential backoff with jitter is mandatory: delay = Math.min(baseDelay * 2^attempt + random(0, 1000), maxDelay).
Lack of connection lost indication. WebSocket doesn't always notify about disconnection (e.g., phone enters a tunnel). Heartbeat solves the problem.
Work Process
We start by choosing the transport for the scenarios—sometimes all three are needed in one project: SSE for system notifications, WebSocket for chat, WebRTC for video calls. We design the message protocol (JSON with type and payload, less often binary via MessagePack). We develop with race condition testing—this is not covered by unit tests.
Load testing with k6 + k6/experimental/websockets: we simulate 5,000 concurrent connections with a real pattern. Our engineers are certified in WebSocket and WebRTC, guaranteeing 99.9% stability.
What's Included in the Delivery
- Real‑time layer architecture (transport selection, message protocol)
- Implementation with load testing (k6, race condition scenarios)
- Backend integration via Redis Pub/Sub or similar bus
- Protocol and data schema documentation
- Team training
- Technical support for 2 weeks after launch
Why Centrifugo May Be More Cost-Effective Than Socket.io?
Socket.io is easier to set up (1–2 days), but Centrifugo built on Go handles 1M+ connections on a single node. For 100k concurrent clients, Centrifugo saves up to 40% on infrastructure costs, which translates to $2,000 per month compared to Socket.io. Get a consultation—we'll help you choose the stack for your load.
Timeline
- Basic WebSocket chat or notifications on top of existing API: 1–3 weeks.
- Collaborative editor with Yjs and persistence: 4–8 weeks.
- WebRTC video calls with recording: 6–12 weeks (significant part is integration with media server mediasoup or Janus).
Contact us to evaluate your project. Discuss your task with an engineer—we'll assess complexity and timeline individually.