Availability
- ≥ 99.9% monthly uptime for the core platform and APIs.
- Service credits applied automatically when a month falls below target.
- Scheduled maintenance is announced in advance and excluded from the measurement.
Enterprise · service levels
Availability you can put in a contract.
We target ≥ 99.9% availability for the platform, design every node for resilience under failure, and back it with continuity guarantees you can hold us to — measured, reported, and credited when we miss.
Availability target
≥ 99.9%
Measured monthly across the platform, excluding scheduled maintenance windows announced in advance.
Service credits
Miss the target in a month and you earn credits automatically — no ticket required.
What we commit to
Three commitments, written down.
An SLA is only worth the obligations it creates. Here is exactly what we promise on availability, response, and support — and what happens when we fall short.
Response
- 15-minute acknowledgement for severity-one incidents, around the clock.
- Named contacts and an escalation path agreed before anything breaks.
- A public status page and proactive notifications for any platform-wide event.
Support
- Dedicated channels with the engineers who run your node, not a queue.
- Post-incident reviews shared in writing, with the root cause and the fix.
- Tiered plans so the level of cover matches what the workload is worth to you.
How the node stays up
Resilience by design, not by luck.
Availability is an outcome of architecture. Each node is built to absorb failure at every layer — here is the sequence that keeps it serving.
- Isolation
Your node runs alone
Dedicated, isolated compute means a noisy neighbour or a tenant-specific fault never reaches your workload.
- Redundancy
No single point of failure
Critical components run with redundant instances across availability zones, so any one of them can fail without taking the node down.
- Failover
Automatic recovery
Health checks detect a failing component and shift traffic away from it in seconds — recovery starts before a human is paged.
- Backups
Recoverable state
Continuous backups with point-in-time recovery mean even a worst-case event has a defined, tested path back to a known-good state.
We will not pretend nothing ever breaks. What we guarantee is how fast we detect it, how it fails over, and what you are owed when the target is missed.
Contractual continuity
What the contract guarantees.
Beyond uptime, an enterprise relationship needs guarantees about your data, your access, and your exit. These sit in the agreement, not just the marketing.
| Your concern | Our guarantee |
|---|---|
| Data durability | Continuous backups with point-in-time recovery and a defined recovery point objective. |
| Data residency | Your records stay in the region you choose and are never moved without written agreement. |
| Continuity of service | Notice periods, migration support, and a documented handover if the relationship ends. |
| Exit and portability | A full export of your data in open formats on request — no lock-in, no hostage data. |
Every guarantee above is enforceable in the master agreement and verifiable against our reporting — not a brochure promise.
Ready to formalise it?
Get the SLA on paper.
We will share the full service level agreement, walk through the credits and continuity terms, and tailor the support tier to your workload.