Targets we commit to. Not metrics we invent.
Per-audience response targets, severity classification, and incident-response posture — tied to the operational incident model the platform actually runs on.
What the platform commits to.
Customer-facing API uptime
99.9% monthly target
Planned maintenance
Announced 48 hours in advance via the partner portal and status.escapra.com
Incident communication
Acknowledged within 1 hour for HIGH severity; status updates at least hourly until resolved
Recovery posture
Fail-closed defaults; operational incident model never throws into business flow
Different audiences. Different windows.
| Audience | SLA | Via |
|---|---|---|
| Hotels — demo requests | Within 24 hours | support@escapra.com · phone |
| Hotels — operational incidents (production) | HIGH: 1h · MED: 4h · LOW: 1 business day | support@escapra.com · status page |
| Distribution Partners — access requests | Within 24–48 hours | support@escapra.com |
| Distribution Partners — operational incidents | HIGH: 2h · MED: 1 business day · LOW: 2 business days | support@escapra.com · webhook |
| Connectivity Partners — enquiries | 2–3 business days | Confidential conversation |
| Investors / analysts — briefings | 3–5 business days | support@escapra.com |
| Security / DPA questionnaires | 5 business days | support@escapra.com |
What HIGH, MEDIUM, and LOW mean.
Production booking flow degraded or unavailable. Source-of-truth degradation cascading to confirmation failures.
Single-tenant or single-scope degradation. Reporting or settlement reconciliation delays.
Documentation, integration questions, scope-change requests outside production paths.
Detect. Classify. Respond. Resolve. Learn.
Detect
Operational incidents are recorded on the booking service — deduped by source, category, entity, and correlation id. Signals are never silent.
Classify
Severity is assigned per the table above. Alert routing respects the configured minimum severity and never throws into business flow.
Respond
On-call engages within the SLA acknowledgement window. Manual recovery is guarded and audit-marked; never silent money or PMS mutation.
Resolve
Resolution is recorded; the incident is queryable via the ops API. Recurrence reopens the incident rather than creating a duplicate.
Learn
HIGH-severity incidents receive a written post-incident review shared with affected customers under their existing support agreement.
Support questions?
Support is handled via support@escapra.com. Production incidents are acknowledged within the SLA windows above.