Skip to main content

Voice Metrics

Voice metrics focus on the audio characteristics of the call.
MetricDescriptionWhat to Look For
Talk TimeTotal time the agent was speaking.Unusually long talk times may indicate the agent is being verbose.
User Talk TimeTotal time the user was speaking.Low user talk time may indicate the agent is not giving the user a chance to speak.
Silence DurationTotal silence in the conversation.Long silences can indicate slow responses or user confusion.
Crosstalk DurationTime the agent and user were speaking simultaneously.High crosstalk can indicate the agent is interrupting the user.
SentimentEmotion in the conversation (positive, negative, or neutral).Negative sentiment can indicate user frustration.

Performance Metrics

Performance metrics focus on the overall success of the call.
MetricDescriptionWhat to Look For
Success RatePercentage of calls classified as successful.A drop in success rate is a clear sign something is wrong.
Failed CallsNumber of calls classified as failed.A high number points to systemic issues.
Hallucination FailuresCalls where the agent provided incorrect information.Any hallucination is a critical issue that should be investigated.
Missed Intent RatePercentage of calls where the agent failed to understand the user.A high rate indicates problems with your agent’s NLU model.
DurationTotal length of the call.Unusually long or short calls may warrant investigation.
CostCost of the call, if provided by your voice AI provider.Helps you monitor operational expenses.

Predefined Evals

Evals go deeper than metrics — each one runs a purpose-built AI judge against the transcript to score a specific behavior. Tuner ships with 21 predefined evals across four categories, optimized using patterns from real-world voice AI deployments. You can enable any combination with a single click, or create your own custom evals. Predefined evals library in Tuner

Accuracy

Checks that the agent captured and confirmed information correctly.
EvalDescriptionScoring
Confirmation accuracyFinal confirmation matches the actual completed actionPASS / FAIL
Data capture accuracyDates, numbers, and key details captured correctlyPASS / FAIL
HallucinationNo invented facts or unsupported claimsPASS / FAIL

Actions & Tools

Validates that tools and workflows were invoked correctly.
EvalDescriptionScoring
Structured intake completenessAll required fields collected before workflow completesPASS / FAIL
Tool args mismatchTool called with correct parameters from the conversationPASS / FAIL
Tool call missingRequired tool is actually invoked, not just claimedPASS / FAIL
Tool response mismatchAgent reports what the tool actually returnedPASS / FAIL
Transfer handlingTransfers executed via tool, not just verbally claimedPASS / FAIL

Conversation Quality

Measures how well the agent communicates throughout the call.
EvalDescriptionScoring
Answer relevanceEach response directly addresses the user’s messageSCORE 1–5
ConcisenessResponses are well-sized with no unnecessary fillerSCORE 1–5
Conversation flow and turn takingLogical progression without loops or abrupt shiftsSCORE 1–5
Dead air handlingPauses handled without silence or over-explainingSCORE 1–5
Escalation handlingEscalates to a human when the situation calls for itPASS / FAIL
Missed intentUser’s goal correctly identified and pursuedPASS / FAIL
PolitenessRespectful and courteous tone throughoutSCORE 1–5
Unnecessary repetitionNo restating of information already coveredSCORE 1–5

Safety & Compliance

Ensures the agent stays within required guardrails.
EvalDescriptionScoring
Identity verification complianceIdentity verified before any sensitive info is sharedPASS / FAIL
Manipulation resistanceHolds its ground against pressure or social engineeringPASS / FAIL
Prohibited contentStays within content boundaries, no off-limits topicsPASS / FAIL
Prompt injection resistanceIgnores embedded instructions that try to hijack behaviorPASS / FAIL
Required disclosuresAll mandatory notices and disclaimers deliveredPASS / FAIL

Setting Up Evals

Getting started with evals takes seconds — no prompt engineering required.

Pick from the library

Open the evals panel, browse by category, and add any predefined eval to your agent with one click. Each eval ships with a prompt optimized on real voice AI use cases.

Agents with optimized prompts

Every predefined eval is backed by an AI judge trained on patterns from actual deployments — not generic LLM prompts. This means fewer false positives and scores you can act on immediately.

Create custom evals

Have a behavior specific to your use case? Write your own eval in plain language and Tuner will run it as a first-class judge alongside the predefined ones.

How to Use Pre-defined Metrics

Overview Dashboard

Key metrics are displayed for a high-level summary. See Best Practices for Agent Monitoring.

Call Logs

All metrics are available as columns for sorting and filtering. See Best Practices for Agent Monitoring

Rule Builder

Use metrics as conditions to create custom labels and red flags. See Introduction to Evaluations & Guardrails.

Next Steps

Learn about real-time alerts and how to set them up.

Introduction to Real-Time Alerts

An overview of real-time alerts, their benefits, and how they help you catch issues early.