---
title: "Voice agent monitoring for every real call"
description: "Speko scores every real call against plain-language checks and groups the failures into issues. It tracks success rate and response time, and keeps the recording and transcript of each call."
canonical: "https://speko.com/voice-agent-monitoring"
last-updated: "2026-11-11"
---

> ## Speko page index
> The complete index of every page on this site is at: https://speko.com/llms.txt
> Read it before exploring further. It lists the exact Markdown URL for every canonical HTML page.

# Voice agent monitoring for every real call

Speko scores every real call and groups the failures into issues.

Speko is a voice agent platform that monitors every real call. It scores each call against plain-language checks. It groups the failures into issues with a trend, so you see what broke and how often.

[Get started free](https://platform.speko.ai/sign-in) [Book a demo](https://speko.com/book)

## Every call gets one result

[Monitor real calls](https://docs.speko.ai/agents/monitoring)

- Handled correctly covers voicemail, no answer, a wrong number and a caller who hung up. Success rate is Success plus Handled correctly, divided by scored calls.
- Monitoring never counts test calls or evaluation runs.

Call results

- Success
- Handled correctly
- Agent mistake
- System error

## Checks in plain language

[Checks](https://docs.speko.ai/agents/monitoring#checks)

- Goal checks test what this agent is for, for example: confirm the patient's name before giving results. Speko drafts them from the agent's instructions, and you can add more.

Basic checks, on every agent

- No made-up facts
- Follows its rules
- Answers what was asked
- No repeating loops
- Ends the call properly

Measured from the call record

- Agent never spoke
- Silence of 8 s or more
- Dropped call
- Speech or voice provider failure

## Health, issues and calls

[Issues](https://docs.speko.ai/agents/monitoring#issues)

- Health shows success rate, real calls, median and p95 response time, agent mistakes and system errors over 24 hours, 7 days or 30 days.
- Speko groups failed calls with the same cause into one issue. Each call opens with its recording and transcript. Make test case turns a failed call into a test.

Issue trend

- New
- Rising
- Falling
- Steady

## Limits

- Scores come from an AI judge, except the measured checks read from the call record.
- If the judge cannot decide a check for a call, it marks the check Did not apply. That result never fails the call.
- Speko publishes no uptime SLA; the service is in public preview.

## Notes

- Every fact is from the Speko page linked beside its section, checked November 10, 2026.
- The product panel at the top is an illustration with made-up data.

## Questions

[Get started free](https://platform.speko.ai/sign-in) [Book a demo](https://speko.com/book)

You get $5 in credit when you sign up. Calls then cost $0.09 per minute.

Related

- [Speko Agents](https://speko.com/agents)
- [Voice agent testing with simulated callers](https://speko.com/voice-agent-testing)
- [Voice agent workflow builder](https://speko.com/voice-agent-workflow-builder)
- [Multilingual voice agents](https://speko.com/multilingual-voice-agents)
- [Voice agent API and SDKs](https://speko.com/voice-agent-api)
- [Voice agent platforms compared](https://speko.com/voice-agent-platforms)
- [Vapi alternatives](https://speko.com/vapi-alternatives)
- [Voice agents by business](https://speko.com/solutions)
- [Speko pricing](https://speko.com/pricing)
