Skip to main content
An enterprise SLA is available on request. The standard Keeper and Guardian tiers ship without a contractual SLA; the operational commitments below describe what we aim for. Customers needing a signed SLA should contact legal@brandatlas.pro.
Effective from: (pending — for customers with a signed enterprise SLA) Last updated: 27 May 2026 This Service Level Agreement (“SLA”) describes Brand Atlas’s uptime commitments for customers who have signed an enterprise SLA addendum to their subscription.

1. Uptime commitment

We target 99.9% monthly uptime of the Service across the application, the API, and the documentation site. 99.9% monthly uptime allows for approximately 43.8 minutes of downtime per month.

2. What counts as downtime

Downtime is time during which the Service is materially unavailable to the customer, measured at the customer’s account. Counted as downtime:
  • HTTP 5xx responses from the application or API in sustained patterns.
  • Inability to sign in.
  • Failure to save brand record changes.
  • Failure to render populated sections or Horizons.
Not counted as downtime:
  • Scheduled maintenance announced at least 48 hours in advance.
  • Outages caused by the customer’s actions (misconfiguration of SAML, exhausted Henry provider quota, custom domain DNS errors).
  • Outages caused by upstream services outside our control, where we have taken reasonable steps to mitigate.
  • Force majeure events.

3. AI feature availability

The AI features (Oswald, Henry) depend on upstream AI providers. We do not provide uptime SLAs for AI providers themselves; their availability is governed by their own SLAs. We aim for 99.5% availability of the Brand Atlas integration layer, which is the layer we operate.

4. Service credits

If our monthly uptime falls below 99.9%, customers with an enterprise SLA are entitled to service credits:
Monthly uptimeCredit
99.5% – 99.9%10% of monthly subscription value
99.0% – 99.5%25% of monthly subscription value
< 99.0%50% of monthly subscription value
Credits are applied against the next month’s invoice. Credits do not aggregate beyond 50% per month.

5. Claiming credits

To claim a credit, write to billing@brandatlas.pro within 30 days of the affected month, with:
  • The dates and times of the downtime experienced.
  • Evidence (screenshots, logs, error messages).
We verify against our internal monitoring; approved credits are applied automatically.

6. Measurement

Uptime is measured by our own monitoring infrastructure. When a dedicated public status page is launched, it will serve as the public source; until then, status is communicated by email and on the Status page in these docs. Customer monitoring may show different figures based on their own infrastructure.

7. Scheduled maintenance

Scheduled maintenance is announced at least 48 hours in advance through:
  • The status page.
  • Email to brand owners on Keeper and Guardian tiers.
  • An entry in the Inbox.
Maintenance windows are scheduled outside Dubai business hours (typically 16:00-22:00 GST is avoided) and stated in UTC and GST.

8. SLA scope

This SLA applies to customers who have signed an enterprise SLA addendum. Without a signed SLA, the uptime targets above describe our operational goal but do not create contractual obligations or entitle the customer to credits.

9. Modifications

We may update this SLA. Material changes are announced 30 days in advance to customers under an enterprise SLA. Customers may terminate their subscription if the change is materially adverse.

What changed

  • 27 May 2026: Initial draft. Enterprise SLA available on request.