AivahHelp CenterGo to Aivah
⌘K
No guide matched “”

Try fewer words or contact support for help.

Contact support

Understand Insights

Choose a useful scope, interpret overview signals carefully, and move from summary cards to underlying records.

Who this is forWorkspace owners, team leads, and reviewers responsible for improving an AI employee

You will achieveUse Insights as a directional starting point, select the correct employee and date range, and validate a finding against detailed records.

Before you beginReview the requirements before changing anything.

Last reviewed 10 July 2026Owner: Product EducationCurrent guide
On this page
  1. Before you begin
  2. Steps
  3. 1. Open Insights
  4. 2. Choose the date range
  5. 3. Read the overview as a set of signals
  6. 4. Open the underlying records
  7. 5. Turn one finding into a controlled improvement
  8. Expected result
  9. Status meanings
  10. Usage credit impact
  11. Information stored or shared
  12. Limits and permanent actions
  13. Common problems and recovery
  14. Next step

This guide is for people reviewing how AI employees are being used. You will set a clear scope, treat the overview as a signal rather than a final business report, and open the underlying records before making a change.

Before you begin

Decide:

  • Which AI employee you are reviewing, or whether you need all employees.
  • The date range and business question you are investigating.
  • Which detailed records can confirm or challenge the summary.
  • Who is authorized to review conversations, leads, quizzes, calls, and personal information.

Use a consistent employee and date range when comparing results before and after a change.

Steps

1. Open Insights

Use the top filters to choose all AI employees or one employee.

2. Choose the date range

Select one day, seven days, thirty days, or a custom range. For a custom range, provide both a start date and an end date.

The workspace may use a different time zone from some visitors. Do not compare totals across systems until the relevant time-zone behavior has been confirmed.

3. Read the overview as a set of signals

The overview can show activity, usage credits, conversation volume, resolution or review signals, lead activity, emotional or sentiment signals, common intentions, quiz and call summaries, items that may need review, and a suggested next move.

Use these cards to decide where to investigate. Do not treat an undefined score or percentage as a complete business total.

4. Open the underlying records

Use Chat Log, Leads, Quiz, or Call Logs to inspect the records behind the question. Read examples rather than relying only on summary cards.

5. Turn one finding into a controlled improvement

  1. Group repeated questions or failures.
  2. Identify missing or outdated business knowledge.
  3. Check whether a connected action succeeded.
  4. Update a draft of the instructions or knowledge.
  5. Retest privately.
  6. Compare results after the update using the same useful scope.

Expected result

You have a clearly defined employee and date range, a directional finding from the overview, and detailed records that support or challenge that finding before you change the AI employee.

Status meanings

  • An item that may need review is a prompt to inspect the detailed record, not a final outcome.
  • A summary, score, or percentage has no approved meaning beyond its visible label until its metric definition is published.
  • If a percentage appears outside a sensible range, use the underlying records and report the display problem to Aivah.

Usage credit impact

Insights can display usage-credit activity. This article does not define how credits are calculated, which activities are included, or when the display refreshes. Use the current workspace as the source of truth and do not derive a per-action formula from a summary card.

Information stored or shared

Insights can surface conversations, submitted lead information, quiz attempts, call details, transcripts when available, and usage. Restrict review to authorized people and use the minimum information needed for the business question.

Limits and permanent actions

  • Some overview calculations can use only the first 50 matching records loaded for review.
  • Search, paging, filters, and downloads can operate on the records available in the current view.
  • This article does not define exact metric formulas, refresh timing, retention, or time-zone behavior.
  • Do not use an undefined score alone for staffing, compensation, eligibility, financial, health, legal, or other high-stakes decisions.

Common problems and recovery

ProblemWhat to try
A summary looks unexpectedly high or lowRecheck the employee and date filters, then inspect the detailed records.
A percentage is outside a sensible rangeDo not rely on it. Use the underlying records and report the display problem.
A result is missingAdjust the employee, date range, or page, then check the relevant detail tab.
Two periods cannot be compared cleanlyMatch the employee, date range length, and available record scope before comparing.
A card's meaning is unclearDo not invent a definition. Record the card label and request the approved metric definition.

Next step

Review conversations, leads, quizzes, calls, and download scope, then use the evidence to test and improve the AI employee.

Keep going