Persistence
Skip to content
Monitor · Preview

Call review

ConversationAppointment request
ActionBooking lookup
OutcomeFollow-up needed
See what happened—and what needs attention.

Monitor the calls.
Understand the outcomes.

See what happened—and what needs attention.

Review conversations, action events and outcomes. Use live oversight and historical evidence to find where the experience needs to improve.

Talk to sales

Monitor capabilities.

Analytics

Review agent activity and outcome measures. Spot patterns that need investigation. Review trends by agent and period, then inspect the calls behind unusual changes.

For regular operational reviews.

How it works +
  1. Choose outcome measures that reflect the task, such as completed requests and escalations.
  2. Review trends by agent and period, then inspect the calls behind unusual changes.
  3. Compare operational measures together so shorter calls do not hide unresolved requests.

Live supervision

Observe an active conversation. Support intervention when appropriate. Give the responsible team access and define when an active call needs intervention.

For calls that need real-time oversight.

How it works +
  1. Confirm which supervision controls are available for your deployment.
  2. Give the responsible team access and define when an active call needs intervention.
  3. Review the conversation and record the result of any intervention for follow-up.

Conversation records

Review logs, transcripts and recordings. Trace the sequence behind an outcome. Compare the transcript and recording with tool events to reconstruct the sequence.

For investigating a specific call.

How it works +
  1. Locate the call using its identifier, time, or available account filters.
  2. Compare the transcript and recording with tool events to reconstruct the sequence.
  3. Apply access and retention rules when sharing records or investigating customer requests.

Improvement suggestions

Review recommendations about agent behavior. Focus attention on recurring weak points. Make a focused change to the prompt, knowledge, or action configuration.

For a cycle of review and refinement.

How it works +
  1. Review recurring issues and the conversations supporting each suggestion.
  2. Make a focused change to the prompt, knowledge, or action configuration.
  3. Re-run relevant scenarios and compare outcomes before releasing the update.

Know where
to look next.

Operations reviews

Understand recurring outcomes across the agent.

A reported issue

Find the call and trace the relevant events.

Post-release checks

Inspect calls after a prompt or workflow change.

Turn call records
into a review process.

Define useful outcomes and the signs of a poor experience for your agent’s task.

Monitor / ConfigurationPREVIEW

Define the essentials.

Setting 01Task completion
Setting 02Unresolved requests
Setting 03Escalation quality
A clear scope for each step
Explore documentation →

Use the call record to understand the conversation and the actions taken.

Monitor / LogicPREVIEW
STARTRelevant transcript
PROCESSAction events
EXPECTED RESULT
Continue
NEEDS ATTENTION
Review
Explore documentation →

Prioritize a specific issue, update the agent and verify the change in testing.

Monitor / ReviewPREVIEW

Review the complete path.

Check 01Review the pattern
Check 02Make a targeted edit
Check 03Retest the scenario
Verify before approving the result
Explore documentation →

Monitor.
Your questions, answered.

Talk to our team
Which metrics show whether the agent is delivering business value?+

Start with the intended outcome: a confirmed booking, a completed enquiry, or a useful handoff. Track completion, repeat contacts, failures, and time to resolution alongside call volume and duration.

How do we distinguish a resolved call from a contained call?+

A call that never reaches a person may still leave the customer without an answer. Review the transcript and downstream result, and define what evidence counts as resolution for each workflow.

What should operations investigate when performance changes?+

Compare affected workflows, agent versions, time periods, and integration results. Sample both successful and failed calls to separate conversation issues from outdated knowledge or unavailable services.

How should we review transcripts and sensitive information?+

Agree on who may access recordings or transcripts and how long they should be retained. Confirm the access and retention controls available for your deployment before using real customer data.

How do we turn call reviews into improvements?+

Group recurring failures, assign an owner, and update the relevant knowledge, prompts, or tools. Add each important failure to a regression test set and compare outcomes after the change.

A person taking a call beside the coast

See the outcome.
Improve the next call.

Monitor prototype

This page is a standalone visual prototype. Interface previews and walkthrough states are illustrative, not live product screens or measured results. No actions, calls or configuration changes are sent by the preview.

Feature scope follows the Persistence feature overview and your supplied website. Verify exact availability, supported operations and configuration before publication. Integration marks are reused from the supplied website, not independently verified connector commitments.

The content structure was informed by the supplied Retell feature-page references. Retell-specific claims have not been transferred to Persistence.