Inference Endpoints: Real-Time Analytics Dashboard with Replica Lifecycle View

AI Impact Summary

Inference Endpoints now expose real-time metrics with a refreshed analytics backend, delivering up-to-the-second visibility into request latency, error rates, and overall throughput for each endpoint. The new Replica Lifecycle View lets operators observe state transitions from initialization to termination, helping diagnose scaling and reliability issues across multiple replicas. These changes accelerate debugging and capacity planning by ensuring dashboards load quickly even at high traffic and by enabling accurate trend analysis through configurable time ranges and auto-refresh.