Blog
Page 5 of 59

LiveKit Agents Observability: OpenTelemetry Traces, Metrics and Audio
What LiveKit Agents observability records per its docs: OpenTelemetry traces, metrics, session reports and recordings, plus what you add yourself.

Voice Agent Call ID Correlation: Joining Twilio, LiveKit and LLM Logs
Fix voice agent call ID correlation across Twilio, LiveKit and LLM logs with one canonical call ID, a mapping table and a join query you can copy.

How to Trace a Voice Call End to End: Phone Network to Agent
Trace a voice call end to end across carrier, telephony provider, transport and agent. Learn where trace context breaks and how to show blind hops.

How to Monitor Voice Agents for Prompt Regressions (With Alert Thresholds)
Set voice agent regression alerts from your own baseline: tag calls by prompt version, slice by intent, gate on minimum calls, and run a 48-hour watch.

How to Root-Cause a Bad Voice Agent Call in 10 Minutes
A minute-by-minute runbook to debug a bad voice agent call: find the moment, read the timeline, check each layer, check changes, then write a verdict.

Voice Agent PII Redaction in Logs and Traces: Where to Mask
Where to mask PII in voice agent logs and traces: source, collector, storage or viewer. Trade-offs, join keys, a 6-step rollout and a seeded-call test.

How to Cluster Voice Agent Failures Into Root Causes
How to cluster voice agent failures: features to group on, rules vs text clustering, ranking by impact, a split test for one cause and a 6-step routine.

Voice Agent Span Attributes: The Fields Every Turn Needs
A per-span field list for voice agent span attributes: call, turn, STT, LLM, tool, TTS and telephony, why each earns its place, and what never to record.

Voice Agent Alerting: What to Page On, What to Ticket and How to Avoid Alert Fatigue
Voice agent alerting for in-house teams: page, ticket, dashboard and log tiers, why each signal sits where it does, and 6 steps to design your alert set.