Deployment & SLA
Continuous delivery across four managed environments, with zero-downtime deployments and automated rollback.
Deployment Model
| Aspect | Details |
|---|---|
| Infrastructure | Managed Kubernetes on Azure (single-region, multi-region on roadmap) |
| Environments | Development, test, staging, and production with automated promotion gates |
| Deployment cadence | Continuous delivery — multiple deployments per week |
| Independent services | Each service builds and deploys independently, limiting the blast radius of any change |
| Zero-downtime | Rolling updates with automated rollback on health check failure |
| Change management | Every change goes through code review, automated testing, and approval gates before reaching production |
Service Level Agreements
99.9%
Uptime Target
Monthly availability target
< 2s
API Response
P95 response time target
< 8hr
RTO
Recovery Time Objective
< 24hr
RPO
Recovery Point Objective
What's Measured
- Included — Core platform APIs and backend services (observability coverage expanding to full stack on near-term roadmap)
- Excluded — Scheduled maintenance windows (communicated 72 hours in advance)
- Measurement — Distributed tracing via OpenTelemetry (SigNoz). External synthetic monitoring and formal uptime reporting are on the near-term roadmap
Production Track Record
Atticus Health powers multi-site care networks — managing over 1 million patient records in production.