Plan requirement
| Subscription | Any plan |
| Also required | Quality Assurance or Workforce Engagement Management |
| Access | Admin |
The settings that apply across every workspace. Most are one-line choices; four of them shape how the whole programme feels.
Where they are
- In Quality Assurance, click your profile icon in the top right.
- Select Settings.
- Open General.
These cover the whole account and every workspace in it, which is the difference between this screen and the workspace settings that look almost the same.
The general settings, and what each decides
| Default workspace | Where new users land. Remember that nobody can be removed from it afterwards. |
| Start of the week | Monday or Sunday, which quietly changes every weekly figure. |
| LLM-based AutoQA | On or off. Four of the eight autoscoring categories exist only when it is on. Switching it off stops the AI scoring but keeps your history and any root causes people added by hand. |
| Automated agent scoring | Turns autoscoring on or off. |
| AI suggestions | Whether you get Enhance suggestions while writing AI prompt-based insights and rating categories. |
| Deep evaluation and validation (EAP) | Shows the reasoning behind AutoQA and Spotlight evaluations in the conversations view. |
| Pass rate (%) | The minimum score for a review to count as a pass. |
| Hide reviewer identities from agents | Reviewer anonymisation, account-wide. |
| Deleted conversations | What happens to reviews attached to tickets that have been deleted. |
The same screen also holds your connections, notifications and scorecards.
The four that matter most
Hide reviewer identities from agents
This is the real switch behind "what agents see". Leave it off unless you have a reason: feedback from a named person is a conversation, feedback from nobody is a verdict.
Pass rate (%)
The minimum score a review needs to pass. Everything anybody says about how the team is doing is measured against this number, so agree it before the first review rather than after the first argument.
There is a harder reason than tidiness: changing this value recalculates every historical review. Reviews that passed last quarter can become failures overnight, without anybody re-reading a single conversation. Move it only deliberately, and tell people before you do.
Automated agent scoring and LLM-based AutoQA
Together they decide how much gets scored without a human looking. Turning them on changes what your reviewers spend their time on, which is the point, but tell the team first.
The connection to Support
What data comes across, and how far back. Check it after any change to your Support configuration, because a new channel does not arrive here automatically.
Disputes are not on this list, and there is no switch for them anywhere. Reviewers and agents can dispute reviews of their own; admins, workspace managers and leads can dispute any review. It is a permission that comes with the role, not something you turn on.
Review them once a quarter
Most were set during setup by somebody making reasonable assumptions about a programme that had not started yet. After a quarter of real use, one or two will be wrong.
Say what changed
Anything affecting what agents see, or how they are scored, should be announced rather than discovered. This is the area where a silent change does the most damage to trust.
Comments
0 comments
Article is closed for comments.