Support reference
Operational reference for support tiers, severities, and escalation. The customer-facing version is Getting started → Support tiers.
Tier matrix
| Tier | Eligibility | Channels | First-response (P1) | Pricing |
| Standard | All accounts | Portal, email | 4 hours, 24/7 | Free |
| Business | SME and above | Portal, email, phone (BWH) | 1 hour, 24/7 | 5% of cloud spend (min BDT 25,000/mo) |
| Enterprise | Enterprise / Regulated FI | Portal, email, phone 24/7, named TAM | 15 minutes, 24/7 | Quote-based |
Severity definitions
| Severity | Definition | Examples |
| P1 | Production down — material business impact | Region unreachable, control plane unresponsive, all VMs in project unavailable |
| P2 | Production impaired — degraded but workable | Slow LB, partial DB replica failure, intermittent IO |
| P3 | Non-prod issue or "how do I…" | Pre-prod environment, configuration question |
| P4 | Feature request, doc bug, low-priority Q&A | "Can the LB do X?", "Doc seems wrong" |
Response targets
| Severity | Standard | Business | Enterprise |
| P1 | 4 hr | 1 hr | 15 min |
| P2 | 8 hr | 4 hr | 1 hr |
| P3 | 1 BWD | 8 BWH | 4 BWH |
| P4 | 3 BWD | 1 BWD | 1 BWD |
BWH = business working hour, BWD = business working day. Asia/Dhaka.
Escalation paths
- P1 stuck for > 1 hr (Standard) / 30 min (Business) / 15 min (Enterprise) → escalate to support manager via "Escalate" button on the ticket
- P1 stuck after support-manager engagement → engineering escalation (Business / Enterprise)
- Enterprise has a direct platform-engineering bridge for P1, available immediately
Communications during incidents
- Status page updated every 30 min during P1
- Customer notification on first response and at every status-page update
- Final RCA published within 5 BWD for any P1