This site is under construction — some content may be incomplete or change without notice.
trailhead
Field Knowledge

FIE Quality

Measure how recommendation outputs are accepted, overridden, and connected to outcomes.

FIE Quality is the operational view of recommendation output health. It reads persisted recommendations and outcomes to show acceptance, overrides, and the available result signals for a selected lodge and time window.

Separate output quality from input quality

QuestionSurface
Are people accepting or overriding recommendations, and why?FIE Quality
Is the underlying lodge, report, seasonal, and client context complete?Context Quality

Low acceptance does not by itself prove that the recommender is wrong. Compare both views before deciding whether to improve the model, the curated context, or the workflow.

Keep the current contract honest

Current v0 recommendations persist ranked beats, score, confidence, rationale, generator metadata, and the feature snapshot used for trace reads. Acceptance or override is recorded during Program selection, and a submitted Fishing Report can close the outcome with actual catch and day quality. Expected-catch delta remains unavailable until that feature exists.

Deeper calibration and periodic FIE reporting belong in the Reports hub; this page is the day-to-day Field Knowledge view.

Next: understand Recommendations and feedback.