Enterprise-grade support parameters, uptime benchmarks, and incident response guidelines.
For enterprise client deployments hosted under our managed cloud setups, we guarantee a **99.9% system availability** SLA. This benchmark is monitored continuously via automated health probes and diagnostic edge logs.
Incidents are classified into severity levels with dedicated response targets. A senior architect is paged automatically on critical severity alerts:
| Severity | Description | Response Limit |
|---|---|---|
| P1 - Critical | Main platform offline or checkout pipelines completely blocked. | < 30 Minutes |
| P2 - Major | Major feature degradation, AI agent alignment drift, or slow API response rates. | < 2 Hours |
| P3 - Minor | Non-blocking UI tweaks, minor content updates, or analytical metrics reporting issues. | < 12 Hours |
SLA retainers include allocated engineering hours per month for preemptive security patches, node package audits, prompt optimization tweaks, and vector database schema adjustments. Unused hours do not roll over but can be prioritized for feature prototyping.
Availability guarantees do not apply to downstream service failures (e.g. global AWS outage, Stripe network breakdown, or upstream OpenAI API outages). Pre-scheduled architectural sprints announced 72 hours in advance are excluded from downtime calculations.
