Evalgent

Blog

Page 2 of 59

Voice Agent On-Call Runbook: The First 15 Minutes of an Incident
Voice AI Testing
13 min read

Voice Agent On-Call Runbook: The First 15 Minutes of an Incident

A voice agent incident runbook for the first 15 minutes: lead, severity by caller impact, scope, containment, status updates and evidence before logs rotate.

November 21, 2026Read more
Turn-Taking Observability: VAD, Endpointing and Barge-In Events
Voice AI Testing
14 min read

Turn-Taking Observability: VAD, Endpointing and Barge-In Events

Which turn-taking events to capture per turn, how to read a turn timeline in a trace, and which findings stay undetermined without the audio.

November 21, 2026Read more
STT Observability: Partial Transcripts, Confidence and Endpointing Events
Voice AI Testing
14 min read

STT Observability: Partial Transcripts, Confidence and Endpointing Events

Which streaming STT events to capture on a span per turn: partials, finals, confidence and endpointing, and what a trace cannot show without the audio.

November 21, 2026Read more
Voice Call Quality Metrics: Jitter, Packet Loss, RTCP and MOS Explained
Voice AI Testing
14 min read

Voice Call Quality Metrics: Jitter, Packet Loss, RTCP and MOS Explained

What jitter, packet loss, RTCP round-trip time and MOS mean, what each can and cannot tell you, and how to attach them to a voice call trace.

November 21, 2026Read more
Voice Agent Telephony Monitoring: SIP Response Codes, Call Setup Time and Post-Dial Delay
Voice AI Testing
14 min read

Voice Agent Telephony Monitoring: SIP Response Codes, Call Setup Time and Post-Dial Delay

SIP response codes voice agent teams should read: 4xx, 5xx and 6xx classes, 486, 487, 503, post-dial delay and setup time, with what to alert on.

November 21, 2026Read more
Voice Agent Provider Monitoring: Detect STT, LLM and TTS Outages Before Callers Do
Voice AI Testing
14 min read

Voice Agent Provider Monitoring: Detect STT, LLM and TTS Outages Before Callers Do

Voice agent provider outage detection: use status pages, per-provider metrics and probe calls to catch STT, LLM and TTS trouble before callers report it.

November 21, 2026Read more
Voice Agent Logs vs What Actually Happened: Where the Agent's Own Account Goes Wrong
Voice AI Testing
14 min read

Voice Agent Logs vs What Actually Happened: Where the Agent's Own Account Goes Wrong

Voice agent logs accuracy has limits: requests logged as results, STT as speech, clocks, dropped events. Check them against carrier and downstream records.

November 21, 2026Read more
Trace Sampling for Voice Agents: Why You Can't Sample Away Failures
Voice AI Testing
15 min read

Trace Sampling for Voice Agents: Why You Can't Sample Away Failures

Voice agent trace sampling: why head sampling drops rare failures on long calls, the miss math, and keep-all rules for errors, slow calls and handoffs.

November 21, 2026Read more
How to Link Call Recordings to Traces: Audio Offsets and Stereo Channels
Voice AI Testing
15 min read

How to Link Call Recordings to Traces: Audio Offsets and Stereo Channels

Link call recording to trace with call and trace IDs, per-leg start times and an offset formula. Mono vs stereo, formats and storage basics for engineers.

November 21, 2026Read more