Service levels
What we commit to on availability, maintenance and support response, by plan, and what we deliberately do not publish yet.
1. Availability
We operate the Service with production discipline: encrypted backups every day, a tested restore drill every week, and health-gated deployments that roll back automatically when a release fails its health check. We do not publish a numeric uptime percentage for self-serve plans. A percentage printed before the history exists to support it is a marketing number, and we would rather say that plainly. Enterprise agreements carry a contractual availability commitment with defined remedies, negotiated per agreement. This section mirrors section 5.1 of the Terms of Service.
2. What counts as available
The Service is available when the core platform is accessible and able to receive and process requests. Partial outages affecting fewer than half of the platform’s features, degraded performance (response times above baseline but below five times the normal average) and intermittent connectivity lasting fewer than five consecutive minutes count as degraded service, not downtime. Degraded service does not qualify for service credits.
3. Scheduled and emergency maintenance
Scheduled maintenance takes place in designated windows, typically between 02:00 and 06:00 UTC on Sundays, with at least 72 hours’ notice by email and in-platform notification. Scheduled windows are excluded from availability calculations. Emergency maintenance to address a critical security vulnerability or a threat to system integrity may be performed with shorter notice.
4. Support response
A security vulnerability reported through the disclosure policy has its own response targets, listed on that page, which are targets rather than contractual commitments.
| Plan | Channel | Response | Who answers |
|---|---|---|---|
| Free | Email and the contact form | Reply within one business day | The engineers who build the platform |
| Solo Pro | Email and the contact form | Reply within one business day | The engineers who build the platform |
| Team | Email and the contact form | Reply within one business day | The engineers who build the platform |
| Enterprise | Email, the contact form and a named contact | A contractual response time, agreed per contract | A named contact plus the engineering team |
5. Service credits (Enterprise agreements)
Where an Enterprise agreement defines service credits, a written claim must reach us within 30 days of the incident with the dates and times of the claimed downtime, a description of the impact, and any relevant logs or screenshots. Credits are applied per the agreement and are not redeemable for cash. Self-serve plans do not carry service credits; the remedy is cancellation under section 4.3 of the Terms.
6. Exclusions
- Scheduled maintenance windows.
- Force majeure events.
- Failures caused by your equipment, software or network connections.
- Actions we take in response to a breach of the Terms.
- Outages of third-party services or infrastructure providers beyond our reasonable control.
- Beta or preview features, which are labelled as such in the product.
7. What we publish instead of a number
- The public status page, readable without an account, so you can check the platform during an incident that affects sign-in.
- Our backup and restore posture, stated on the trust centre and in the annex to the Data Processing Addendum.
- The deployment discipline: every release is health-gated, with the previous image kept for an immediate rollback.
Related documents
Questions about this document go to the contact form with the topic set to legal; a reply comes from a person within one business day.