Architecture
Standards-first. Modular. Built for sovereign and SaaS deployment.
Six layers, cleanly separated. Strong consistency for safety-critical flows. Event-driven everywhere it makes sense. Cloud-neutral, multi-region, and offline-capable at the edge.
Layered topology
The shape of the platform.
Experience layer
- Clinician web & mobile
- Patient portal & app
- Admin console
- Partner / SMART on FHIR apps
API & event plane
- FHIR R4/R5 REST + Bulk Data
- GraphQL composition
- Event streaming (Kafka-compatible)
- Webhooks & subscriptions
Clinical domain services
- Patient & encounter
- Orders & results
- Medication & eMAR
- Clinical documentation
- Revenue cycle
Platform services
- Identity Broker
- Authorization Decision Service
- Consent & policy engine
- Audit ledger
- Terminology service
Data plane
- Operational FHIR store
- Lake / OMOP / mart for analytics
- De-identification & pseudonymisation
- Tenant isolation + CMK
Integration fabric
- HIE — IHE XDS/XCA, FHIR
- HL7 v2, X12, NCPDP
- National exchanges (CIS2, TEFCA, MyHealth@EU, DHIS2)
- Migration factory
Non-functional
High availability, DR, and patient-data resilience.
Multi-region active/active
Stateful clinical writes replicated with strong consistency where it matters; eventually consistent elsewhere with conflict resolution.
Backup + point-in-time restore
Continuous backup of operational stores. Per-tenant restore windows. Cross-region snapshots with customer-managed keys.
Disaster recovery drills
Documented RPO/RTO per deployment. Regular game-day failovers. Restore evidence packaged for hospital and authority audits.
SaaS multi-tenancy
Tenant isolation enforced at data, compute and identity layers. Noisy-neighbour controls and per-tenant observability.
Sovereign + private cloud
Same code, different deployment profile. Regional data planes. EU-only or country-only routing where required.
Edge & offline modes
Edge clinical capture with deterministic sync — critical for emerging-market deployments and field operations.
Migration factory
Seamless data migration from Epic, MEDITECH and others.
A repeatable migration pipeline: source profiling, terminology mapping, FHIR transform, validation, dual-running, and clinician-led cutover.
- Step 1Profile
- Step 2Map
- Step 3Transform
- Step 4Validate
- Step 5Cut over