# How To Monitor Evaluation Runs

## Objective

Track queued, processing, completed, skipped, and failed QA work and open its audit record.

## Before you begin

You need company-scope access to raw QA grades: **Quality: Manage** or **View Only**. **Team Only** and **Own Only** do not open the company-wide run list.

## Resolution

1. **Open the run list.** Go to [QA Analysis · QA Grades](https://wfm.zenops.cx/service/quality). Each row identifies a ticket, category, origin, score, critical result, status, and time.
2. **Narrow the view.** Use the date range, exact ticket ID, agent, category, critical-failure, and score-band controls. Open **Filters** for run origin or pipeline/result status. Select **Refresh** to reload the latest state.
3. **Inspect a row.** Expand a run for a quick outcome and choose its full run page to see the frozen scorecard, target grades, and evidence. A run may have an overall ticket grade, agent grades, or, for an agent batch, only the selected agent's grade.
4. **Follow a batch.** Open [Grade Batches](https://wfm.zenops.cx/service/quality/grade-batches) to see selection at launch, progress counts, and each ticket. Its **View all runs** link shows the batch's runs, including retries; the usual date filter does not narrow that batch view.
5. **Investigate exceptions.** Distinguish waiting or processing work from a terminal skip or failure. Open the run for its stated reason before changing policy or retrying. Use [Investigate A Skipped Or Failed Evaluation](/help/articles/qa-investigate-a-skipped-or-failed-evaluation/) for a focused check.

Completed results are AI drafts until a manager reviews and publishes feedback. A completed run alone does not make a result visible in My Quality.

> Canonical page: https://zenops.cx/help/articles/qa-how-to-monitor-evaluation-runs/
> Audience: For managers and analysts
> Summary: Track QA run progress, results, skips, and failures.

## Visual references

- The Grade Batches list shows batch type, agent or category, solved dates, ticket count, status, progress, average weighted score, and start time. Agent names are generalized in this example.: [QA Analysis Grade Batches list with three completed agent batches, generalized labels Agent 1, Agent 2, and Agent 3, solved dates, ticket counts, progress, weighted scores, and start times.](https://zenops.cx/help/assets/qa-grade-batches-anonymized.png)
- A batch detail page summarizes completion and outcomes, then lists the tickets with their categories, scorecard versions, agent grades, scores, critical results, and links to each run. The selected agent name is generalized in this example.: [QA Analysis detail page for a completed agent batch labeled Agent 1, showing date range, category count, run duration, completion progress, outcome totals, average weighted score, and a ticket table with run links.](https://zenops.cx/help/assets/qa-grade-batch-details-anonymized.png)