LiveKit Scaling EXPERT for 3,000+ Concurrent Calls

Our production LiveKit deployment requires scaling to support at least 3,000 concurrent audio calls. Seeking an expert with proven experience in Kubernetes clusters, autoscaling, TURN servers, SFU optimization, distributed Redis, and robust observability for high-scale environments.

Engagement scope:
• Audit current architecture via screen share, demonstrating prior projects (best if scaled beyond 3,000 sessions)

• Implement scaling: node grouping, bandwidth management, session migration, stress testing, and alerting.


• Ongoing: proactive maintenance, performance optimization, incident response, and capacity reviews.

Success metrics:
1 Soak test validating 3,000+ participants with <400 ms latency and zero dropped packets.
2 Resource utilization (CPU, memory, network) within agreed thresholds.
3 Grafana dashboards and alerts for preemptive issue detection.
4 Repeatable deployment scripts (Helm/Terraform equivalent) committed to our repository.

Hourly engagement with time-tracking. Must be proficient in LiveKit internals and large-scale SFU deployments; no learning curve tolerated.