Retention and deletion

The lifecycle job implements purging for expired analytics and behavioural aggregates. Actual deletion depends on successful job execution; this setting does not establish provider or backup retention.

The lifecycle job implements deletion of expired Reader feedback and optional comments after the configured period. Successful execution and any residual copies require separate verification.

Direct support messages remain available to the participating customer and authorised Beetl platform staff. They are removed when the participant membership or organisation is deleted; a shorter configurable support-message period is not yet implemented.

The learning retention setting defaults to 365 days for supported records. Legacy learning progress is retained pending verified migration; assessment responses, uploaded evidence and acknowledgements use separate retention rules. This setting does not guarantee deletion of every learning record after 365 days.

The lifecycle job implements deletion of expired stored assistant questions and source references. Generated drafts and provider-held records have separate lifecycles; this is not a guarantee that every AI record is erased after 90 days.

The signed download grant expires and owner membership is checked again. The generated JSON is not persisted by Beetl.

Confirmed owner deletion requests become eligible after the cooling-off period. The lifecycle job removes eligible tenant database records transactionally and queues private-object cleanup; failed cleanup stays queued for retry. Global account records, lawful retention exceptions, provider copies and backups require separate handling. Workspace deletion does not prove erasure of every copy.

Backup retention and deletion propagation depend on the production provider. The approved customer agreement must state the period and restore isolation procedure.