AutoQA Dashboard

How to interpret the results of AutoQA auto scoring

Dajana avatar
Written by Dajana
Updated this week

The Auto QA Dashboard provides information about auto-reviewed conversations, manually reviewed conversations, auto-reviewed agents, and manually reviewed agents.

  • Auto-reviewed conversations - Number of conversations that had at least one category automatically scored, without human intervention.

  • Manually reviewed conversations - Number of conversations that have been manually done by human reviewers.

  • Efficiency gain - It displays how many times more can AutoQA cover tickets compared to manual reviewing.

  • Auto-reviewed per reviewee - The average number of reviews by AutoQA for each reviewee.

  • Manually reviewed per reviewee - The average number of reviews manually given to each reviewee.

  • Acceptance rate - Percentage of AutoQA scores that are accepted by users as accurate.

  • Modified AutoQA conversations - Number of conversations where human has changed the AutoQA score in at least one category.

  • Breakdown for top languages - Ratio and count of conversation languages.

  • Category insights - Breakdown of AutoQA scores per categories.

  • Category scores over time

  • Categories per reviewee - Information on reviewees and their automatically calculated average score of specific auto category for selected time period.

  • Root causes of spelling and grammar - Information about reviewees and the automatic root causes from the auto category spelling & grammar.


Filters:

Time filter - Displays results based on the time frame chosen.

Workspace filter - Displays results based on workspaces selected, which have any Auto QA review results.

Reviewee filter - Displays results based on the reviewees selected.

Rating category filter - Displays results based on auto categories selected.

Language filter - Displays results based on conversation languages.

Root cause filter - Displays results based on available root causes (only grammar root causes available at the moment).

Sentiment filter - Displays results based on customer sentiment.

Ticket channel filter - Displays results based on set channels.

Outliers filter - Displays results based on Outliers.

Connections filter - Displays results based on set connections.

FAQ

Why am I seeing more than one autoQA category in the filter?

Currently archived and deleted categories are also shown in the category filter.

If I manually change a rating that was auto scored as N/A does that affect the acceptance rate?

Acceptance rate takes into account any changes to given auto score when submitting, meaning N/A as well.



Did this answer your question?