Appearance
Datapoint Retention
Datapoint Retention controls how long YuzeData keeps the datapoints your workflows produce. Without a policy, datapoints are kept indefinitely.
Navigate to Company → Datapoint retention.
How a policy works
A policy pairs a scope with a retention period.
| Scope | Applies to |
|---|---|
| All schemas | Every schema without a policy of its own. Shown as the default at the top of the page. |
| One schema | Every bucket of that schema, except a bucket with its own policy. |
| One bucket of one schema | Just that bucket. |
The most specific policy wins: a schema set to delete after 90 days with one bucket set to keep indefinitely deletes everything in that schema except that bucket. Each scope holds one policy — YuzeData names the existing one if you try to add a second.
Periods count from when the datapoint was created in YuzeData, not from the business date it carries. A record about a three-year-old incident that arrived today is one day old.
Saving does not delete
YuzeData removes expired datapoints on a schedule, roughly every eight hours, so they stay visible until the next run. Before you save, the dialog shows how many datapoints the policy covers, how many are already past the period, and how many are kept.
Buckets a workflow reads back need their own policy
Check this before setting a broad period.
Some steps read their own output back as memory — deduplication compares incoming records against what it wrote previously to decide what is new. Delete that history and the next run treats everything as new, re-sending records the receiving system already has.
WARNING
Give such a bucket its own policy set to Keep indefinitely; it overrides the schema's. To find them, look for workflow steps that read from the bucket they produce into.
The dialog says when a bucket holds no datapoints yet. That is fine: you can write a policy before the first run, and the bucket field accepts any name you type.
Removing a policy widens deletion
Datapoints fall back to the next broader policy. Remove a keep indefinitely on a bucket whose schema deletes after a period, and that bucket starts being deleted on the next run.