Disaster Recovery as a Service
Always ready to recover. Pay only when you activate.
Fully managed DR on United Private Cloud®. Replicated around the clock, tested by AI, billed only when you drill or declare.

Recovery time, Platinum tier

0 min

Savings vs. hyperscale DR

0 %+

Availability,
N+M

0 %

Managed and monitored

/7
Included
How it works

Replicate. Drill. Fail
over in minutes.

Three steps, fully managed.
1 · Replicate

Copied in near real time

VMware ESXi, Hyper-V and KVM, to a recovery site in
your jurisdiction.
2 · Drill
AI proves it works
Drills on an isolated copy, with
a report for auditors.
3 · Fail over and back
Running again in minutes
Declare, recover, fail back in order. Pay only while
compute runs.
Why it costs less

Stop paying for a second
site you may never use.

Warm replication. Committed targets. Compute billed only when it runs.
Always ready, never idle

Warm replication, committed targets

Continuous replication to a second data center. RPO and RTO committed in your subscription.

  • Near real-time replication
  • Committed RPO and RTO
  • Drills prove it works
Warm replication, committed targets
Every day before the disaster

Recovery that tests itself

Three AI capabilities, working in the background.

AI-driven DR drills

Drills run on an isolated copy. Recovery is proven without touching production.

Proven recoveryNo downtimeReport for auditors
When the disaster is an attack

Production can be encrypted.
Your recovery copies cannot.

Three steps, fully managed.
Immutable by default
Copies attackers cannot delete
Locked until retention ends, away from your network.
Early warning
Encryption spotted early
AI marks the last clean recovery point for you.
Clean-room recovery
Come back clean
Recover in isolation, verify, then return users.
Beyond the basics

Disaster recovery for
AI, for ransomware and
for your next platform

Pick one.

DR for your enterprise AI stack

Models and data come back with everything else.

  • Model registries and data pipelines
  • Training and inference data
  • AI delivery continuity
ModelsDataPipelines
DR for your enterprise AI stack
See it in motion

Watch DR at work

Six moments from UPC DRaaS, from replication to failover and back.
Redundant paths. Lose a network path and traffic moves to the other automatically. Ransomware meets immutable backups. Production is encrypted, but the backups cannot be deleted. The right way back for each disruption. A fire, a cyberattack or a network cut each get the recovery path that fits. Pay only when you activate. Storage every month, compute only for a drill or a real activation. Your site and ours, linked. Your data center on one side, your recovery site on the other, kept in sync. DR drills that prove it works. Production replicates to the DR site, and a drill proves recovery.
LIVE · AUTO-PLAYING
Your promise

Three committed
tiers. One price model.

Each tier commits to data loss and downtime. All bill compute only when used.
Platinum
Near-zero

data loss

15 min

back online

Core databases, ERP, payments.
Gold
< 15 min

data loss

1 hour

back online

Customer-facing and line-of-business apps.
Silver
< 1 hour

data loss

4 hours

back online

File servers, internal tools, test and dev.
Sign-off

Recovery your regulator
and your board can sign off

Secure, compliant, sovereign and cheaper. Pick a proof point.
Secure

Recovery stays on isolated networks

Every rule is checked against policy. An open port is blocked before anyone can use it.

  • Isolated recovery networks
  • Rules checked continuously
  • Risky change caught in seconds
Recovery stays on isolated networks
↑