Emplyr

Support and service policies

Backups and recovery policy

Last updated: 21 October 2026

Application database

The shared production Emplyr application database currently uses Azure SQL automated backups with a configured 7-day point-in-time recovery window and locally redundant backup storage. Long-term backup retention is not enabled. Recovery depends on Azure backup availability and the available restore points; this is not a guarantee of zero data loss or recovery within a fixed time.

These backups cover data held in that database. They do not automatically cover separate Process Hub runtime databases, Fabric workspaces, external data sources or files held outside it.

Application file storage

The production application Blob storage account currently has 7-day blob and container soft deletion enabled. This provides recovery for supported deletion scenarios within that window. Versioning and Blob point-in-time restore are not enabled.

Soft deletion is an in-account recovery feature, not an independent backup. It does not provide a complete recovery solution for storage-account deletion, a regional disaster or every modification scenario.

Fabric and Process Hub data

For BYOF, the customer or its service provider is responsible for protection and recovery of customer-managed Fabric items, runtime databases and source data. For Managed, Emplyr coordinates recovery of resources it manages, subject to the recovery features available for each resource. A Fabric capacity reservation is a compute commitment, not a backup service.

Do not assume that every report, semantic model, lakehouse, workflow record or uploaded document is covered by the application database backup. Any additional backup scope, recovery objective or archive requirement must be agreed and verified for the relevant environment before relying on it.

Requesting recovery

Contact support promptly with the organisation, instance, affected data, time of the incident and last known correct state. Avoid further changes to affected data where practical. We assess available recovery options and obtain the appropriate customer authorisation before restoring customer content.

Shared-database recovery must preserve tenant isolation. Restoring an entire database is not a promise of individual-record recovery. Recovery time and recoverable data depend on the incident, resource and available restore points. No fixed recovery-time or maximum-data-loss guarantee is made by this policy.

Let’s start with your context

What would better
look like for you?

Tell us where you are starting. The right conversation is about your environment and the outcomes you need.