2026-11-06 - Tenancy Health: Sharper Rankings and New Event Checks¶
Release¶
AgileData.io - Fewer False Alarms: Tenancy Health Learns From Real Tenancies
Running Tenancy Health against real tenancies showed where findings were ranked too high, too low, or raised when nothing was wrong. This update fixes those cases and adds new checks for event tables.
BigQuery ML models no longer show up as missing tables. Rules with placeholder names are recognised more accurately. A tap whose schedule has disappeared is now reported as missing once the check has confirmed where your schedules live, instead of as something the check could not see. Grain findings on a table that is also listed as an orphan are ranked as tidy, so the noise doesn’t drown out real problems.
Event tables now get their own grain checks: duplicate copies of the same event, participants that don’t match any concept, and how consumed rows compare with the event’s current rows.
What’s New:
BigQuery ML models no longer raise false findings
More accurate detection of placeholder rule names
Missing tap schedules reported clearly once their location is confirmed
Grain findings on orphan tables ranked as tidy
New grain checks for event tables
Rules built with an older parser are named when they build a failing table
What this means for you:
A cleaner findings list you can trust
Problems with event tables caught as well as with concepts and details
Less time spent ruling out findings that weren’t real
Last Refreshed¶
Doc Refreshed: 2026-11-14