These are illustrative scenarios showing how the products are used, not customer case studies.
The same sequence, twice
Today
- Auditor asks about a date
- Only the current row exists
- Correlate logs and exports
- Answer is an estimate
With SynaptaGrid
- History on for the schema
- A snapshot per write
- Read that point in time
- Restore, itself recorded
The situation
The current row is the only state the database keeps. Reconstructing an earlier one means correlating application logs, a nightly export and somebody’s memory of a schema change — expensive to do once, and not repeatable in front of an auditor.
How it is approached
History is enabled per schema, and every write keeps a snapshot to a depth chosen for that schema. Any retained point can be restored, and the restore is itself recorded, so using the history never rewrites it. Model versions are kept alongside the data, and every change is published to the event stream in the same transaction as the change, carrying a field-level difference.
What changes
The state of a record on a date becomes a lookup rather than a reconstruction. Because the event is written in the same transaction as the change, what downstream systems recorded and what the database holds cannot quietly disagree.