Service Level Agreement (SLA)
Effective Date: March 23, 2026
This Service Level Agreement ("SLA") is part of the Terms of Service between Kosmatic Inc. ("Kosmatic," "we") and you ("Customer"). It defines our uptime commitment, how we measure it, and your remedies if we fall short.
1. Uptime Commitment
Kosmatic commits to 99.9% monthly uptime for the core Service, measured as:
Uptime % = ((Total Minutes in Month - Downtime Minutes) / Total Minutes in Month) x 100
"Downtime" means the Service is materially unavailable to Customers, as measured by our server-side monitoring (Render health checks + external synthetic monitoring). Partial degradation (e.g., slow dashboard load, delayed analytics) does not count as Downtime unless core API endpoints return errors for >50% of requests over a 5-minute window.
2. SLA Credits
If monthly uptime falls below 99.9%, you may request a credit on your next invoice:
| Monthly Uptime | Credit (% of Monthly Fee) |
|---|---|
| 99.0% -- 99.9% | 10% |
| 95.0% -- 99.0% | 25% |
| Below 95.0% | 50% |
- Credits are not cumulative -- maximum 50% credit per month
- Credits are applied to future invoices only; no cash refunds
- You must request credits within 30 days of the affected month by emailing support@kosmatic.com
- Credits are your sole and exclusive remedy for downtime
3. Exclusions
The following are not counted as Downtime and do not qualify for SLA credits:
- Scheduled maintenance -- We provide 24 hours' notice via email and dashboard banner. Maximum 4 hours per month. Typically performed during low-traffic windows (02:00-06:00 UTC).
- Emergency maintenance -- Security patches or critical fixes that cannot wait for a maintenance window. We provide notice as soon as reasonably possible.
- Third-party service failures -- Downtime caused by Shopify, Klaviyo, HubSpot, Meta, Google, Render infrastructure, or other third-party providers outside our control.
- Force majeure -- Natural disasters, wars, pandemics, government actions, power outages, internet backbone failures.
- Customer-caused issues -- Misconfiguration, excessive API usage beyond rate limits, DDoS attacks originating from Customer infrastructure.
- Beta/preview features -- Features explicitly labeled as beta, preview, or experimental.
4. Maintenance Windows
| Type | Notice | Duration | Frequency |
|---|---|---|---|
| Scheduled | 24 hours (email + dashboard) | Max 4 hours/month | As needed (typically monthly) |
| Emergency | Best effort | As needed | Rare |
5. Support Response Times
| Severity | Definition | Initial Response | Update Frequency |
|---|---|---|---|
| Critical (P1) | Service completely unavailable; data loss or security breach | 1 hour | Every 2 hours |
| High (P2) | Major feature unavailable; significant performance degradation | 4 hours | Every 8 hours |
| Medium (P3) | Non-critical feature issue; workaround available | 1 business day | Every 2 business days |
| Low (P4) | Cosmetic issue, feature request, general question | 2 business days | As needed |
Support hours: Monday-Friday, 9:00 AM - 6:00 PM Pacific Time (excluding Canadian statutory holidays). Critical (P1) issues are monitored 24/7 via automated alerting.
6. Incident Communication
During service disruptions:
- We post real-time updates to our status page
- Email notification sent to affected Customers within 30 minutes of confirmed incident
- Post-incident report (root cause + remediation) provided within 5 business days for P1/P2 incidents
7. Data Durability
Customer data stored in our PostgreSQL database is:
- Encrypted at rest (AES-256)
- Backed up daily with 7-day retention
- Hosted on Render's managed infrastructure with automatic failover
We do not guarantee against data loss caused by Customer actions (e.g., using the "Delete Profile" or "Wipe Visitors" features).
8. Platform-Specific SLAs
| Component | Target | Measurement |
|---|---|---|
| Dashboard & API | 99.9% uptime | Server-side HTTP monitoring |
| Tracker Script | 99.9% uptime | CDN availability + endpoint health |
| Webhook Processing | 99.5% delivery | Webhook delivery logs |
| Integration Sync | Best effort | Dependent on third-party availability |
| AI Workflows | 99.5% execution | Workflow run logs |
| MCP API | 99.9% uptime | API endpoint monitoring |
9. Changes to This SLA
We may update this SLA with 30 days' notice. Material changes that reduce our commitments will be communicated via email. The SLA in effect at the time of an incident governs credit eligibility.
10. Contact
To report an incident or request SLA credits:
Email: support@kosmatic.com
Subject: "SLA Credit Request -- [Month/Year]"