Waypoint Claims
Trust / Backups and recovery

A backup nobody has restored is a hope

An encrypted database backup every hour, write-locked for 30 days and copied to an isolated account in a second US region. A standby in that region. A restore target of 4 hours, with at most 1 hour of work lost — tested every week.

With continuous replication to the standby region, a region failure loses seconds. The 4-hour, 1-hour figures are the floor for restoring from backup.

Ask for the DR position How it works
The question this answers
How do you know your backups work?
Because we restore them every Sunday and time it. The weekly test restores the newest backup into a throwaway database and records the time. Every six months a recovery review measures what we achieved against the targets, and every year an independent drill restores closed records from the cold vault without the live application.

How copies are held and proven

Hourly, isolated, drilled
01
Hourly, With Point-in-time Recovery
An encrypted backup every hour, kept for exactly 30 days and write-locked for the same 30 days — no longer, no shorter.
02
An Isolated Backup Account
Backups are copied to a separate account in the second US region that normal production logins cannot reach, so a stolen production login cannot alter or delete them.
03
A Standby Region
The managed database replicates continuously to the standby. A separate read-only copy answers reports, so heavy reports never slow claim work.
04
A Person Switches, By Checklist
On an alarm, a person switches regions by the checklist; in-region failover is the managed database’s own. The public address moves within a 5-minute DNS lifetime.
05
Tested Weekly, Reviewed, Drilled
A Sunday restore test, a six-monthly recovery review against the targets, and an annual cold-vault archive drill.
The DR position

Targets, and how they are measured

Waypoint’s Platform Health card shows the last restore’s elapsed time and the projected restore time at today’s size, read off the system rather than estimated.

Hosting
One United States region of a major cloud provider, with a standby in a second US region
Recovery Time Target
Restore within 4 hours
Recovery Point Target
At most 1 hour of work lost; with replication, a region failure loses seconds
Uptime Target
99.9%, measured every minute by an outside monitor, with an alarm when the standby stops copying
Backup Keys
Backups of your data are locked under your key. Waypoint’s backup logins are kept apart from production logins.
Documents
Live documents, backups and archives copied to the second region and the isolated account
After A Restore
The disposal engine runs at once, so disposed records never come back
Where we are honest

What a region switch still asks of you

We would rather you know the shape of a switch now than discover it during an incident.

A Person Makes The Switch
Region switching is a checklist a person runs on an alarm, not an automatic cutover. In-region failover is automatic; cross-region is deliberate.
People Sign In Again
After a switch, users sign in again. Queued work is retried from the database’s own record.
Targets Are Targets
Four hours and one hour are what we design and drill for. Ask us for the latest measured restore, not just the target.

Read next

Audit & Retention The warm store and the cold vault behind the backups. Monitoring The alarms that start a switch. Operator Portal The Platform Health card, where recovery is measured.

Ask for the drill figure

Not our target — the time last Sunday’s restore actually took, and the projection at today’s size. Any vendor who cannot answer that has not restored anything lately.

Book a walkthrough Back to the overview