Engineering zero-downtime, multi-region cloud architectures for medical device software operating in critical clinical care pathways.
When medical software guides intensive care therapy or acute diagnostic triage, system downtime translates directly into compromised patient outcomes. Life-critical SaMD requires an RTO under 60 seconds and an RPO of zero (zero data loss).
Architectural blueprints for active-active database clustering across geographically separated cloud availability zones, using CockroachDB or Amazon Aurora Global Database with automated DNS failover (Route 53) to survive catastrophic regional datacenter outages.
Implementing automated chaos engineering experiments (Chaos Mesh): deliberately terminating cloud instances, injecting network latency, and severing database links to mathematically verify system self-healing capabilities.