The following feature flags and permissions are required to use this feature:

Required Permissions:

Evaluation quotas set how many evaluations each agent should receive in a given period and how that work is split among your evaluators. Set them once and evaluagent does the heavy lifting — pulling contacts into queues and distributing them according to your rules.

Go to Evaluate > click Auto work-queues.

You'll see your existing work queue templates. Quotas are configured per template, so either open an existing template or create a new one.

The replenishment cycle decides how often quotas reset. Edit the template's General settings and choose one of:

Pick the cycle that matches how your team plans QA work, not how often agents are scored.

Go to the Contact Types tab on the template and select (or create) the contact type you want to set a quota on.

A contact type defines which conversations land in the queue — for example, "voice calls from the sales team" or "support emails over 5 minutes long".

Find the Set your target number of evaluations per agent field and enter how many evaluations each agent should receive each cycle.

evaluagent calculates the total evaluations automatically:

Total evaluations = number of agents × evaluations per agent

For example, 10 agents at 5 evaluations per agent gives 50 evaluations per cycle.

If you're using an automated scorecard and don't want a fixed quota, you can choose I want to evaluate all contacts that match this filter instead. evaluagent will show you the number of contacts available in the last 7 days so you can sanity-check the volume.

Scroll to Add your evaluators and add the people who'll work the queue. For each evaluator, set their share of the total evaluations as a percentage.

Rules to know:

evaluagent works out each evaluator's quota for you:

Evaluator quota = total evaluations × (share ÷ 100)

So in a queue of 50 evaluations:

If you're using an automated scorecard with human oversight, use the slider to set what percentage of AI-scored evaluations need a human to review them. Anything from 0% to 100%.

If you want every AI evaluation reviewed, set it to 100%. If you trust the AI on most contacts and just want spot-checks, set it lower.

Click Save to apply the quota.

The next replenishment cycle will populate the queue based on your settings.

Once the queue is live, you can track how the team's doing.

Go to Evaluate > Auto work-queues and click the Progress button. The progress view breaks down completion rates by:

The work queue list page shows an alert if any active agents aren't covered by an active work queue. That's your prompt to extend a template or create a new one — agents not in any queue won't get evaluated.