Illustrative sample

Voice Agent Evaluation Report

A representative view of the evidence Vocalyze produces. Names and results below are fictional and do not represent a customer.

Overall score

76 / 100

Needs review

Goal completion

Partial

Escalation incomplete

p95 latency

2.8s

Target: 1.5s

Language

English

en-US caller

Rubric breakdown

Where the call passed and failed

Identity verified before account disclosure

Agent confirmed two required identifiers.

100

Issue resolved or correctly escalated

Escalation started, but the promised handoff confirmation was missing.

62

No unsupported claim

All policy statements matched the supplied knowledge.

100

Response latency under 1.5 seconds

Two turns exceeded the target; p95 was 2.8 seconds.

54

Transcript evidence

Simulated caller

The last payment still appears twice. Can you connect me with someone who can correct it?

Voice AI agent

I can escalate this to our billing team now.

1.1s
Simulated caller

Yes, please. Will they receive the details from this call?

Voice AI agent

Thank you for calling. Goodbye.

2.8s

See this level of evidence on your own agent.

Start with one important workflow and turn the findings into a repeatable regression suite.

Request a reliability audit