> For the complete documentation index, see [llms.txt](https://doc.duaer.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://doc.duaer.com/advanced-ai/evaluations/metric-based-evaluations.md).

# Use metrics to measure quality in Duaer

In Duaer: Run an AI digital organization against a test set and record scores you calculated, so you can compare versions over time.
## Test set and Evaluation node

Use Evaluation Trigger (or the test-data source the docs name) to take one test row at a time. After the digital organization runs, the Evaluation node collects scores. Scores show on the evaluations view. They do not have to be written back to a sheet.

## Calculate metrics <a href="#2-calculate-metrics" id="2-calculate-metrics"></a>

1. Before the Evaluation node, use Code, comparison nodes, or Set Metrics to produce fields such as accuracy or relevance.
2. Map those fields into the Evaluation node.
3. Open the evaluations view, run a test, and compare the summary scores.

The “view metric examples” note on the node links to this section. Mapped names must match the fields you calculated earlier. Otherwise the evaluations view is empty or wrong. In production, wrap metric logic behind a check-if-evaluating branch when you do not want to score every live run.
## Questions

### What does a metric-based evaluation do in Duaer?

It runs the digital organization against a test set and records scores you calculated. Scores show on the evaluations view. Calculate metrics before the Evaluation node, then map them in.

### Where do evaluation metrics in Duaer come from?

From fields you calculate in the digital organization, or from Set Metrics on the Evaluation node. Duaer does not invent a score that has nothing to do with your work.