Deployment Architecture Diagram

Infrastructure layout showing how CMS is deployed as independently deployable frontend and backend components, protected through Zscaler, Akamai, and Azure Application Gateway, with private AKS runtime and separated operational and analytical data stores
Authorized Riyadh Air Users (Browser Access)
USER ACCESS & EDGE ENTRY LAYER
Controlled user access path and public edge protection before Azure origin entry
Zscaler
Zscaler
Enterprise-controlled user access layer enforcing zero-trust access and approved enterprise access paths
Akamai
Akamai Edge
Primary public entry point providing WAF, DDoS protection, bot protection, and controlled edge routing
Azure Application Gateway
Azure Application Gateway (WAF)
Controlled origin entry, L7 routing, inspection, and protected forwarding to AKS ingress
TLS
Protected Inbound Path
Internet → Zscaler → Akamai → AppGW → AKS Ingress with no direct public AKS exposure
Frontend and backend are deployed separately; the browser never accesses Fabric, DocumentDB, ADLS, or private backend services directly.
FRONTEND DEPLOYMENT LAYER
Separate frontend deployable artifact delivered through the edge layer
React
React SPA
Web-only React single-page application providing dashboards, workflows, alerts, filtering, and interactive user actions
UI
Frontend Static Build
Own deployable artifact with layout assets, branding, icons, and UI resources packaged with the frontend build
CDN
Frontend Delivery via Akamai
Frontend is delivered separately through Akamai rather than deployed as an AKS backend workload
BACKEND COMPUTE LAYER – PRIVATE AKS
Multiple independently deployable backend services/components hosted on a private AKS cluster
ING
AKS Ingress
Internal ingress reachable only from Azure Application Gateway
API
CMS Backend API Services
Frontend-facing secured APIs exposed by backend services with no dedicated BFF layer
DOM
Domain Services
Business capability services such as Roster, Attendance, Events, Trainings, Qualifications, Recruitment, and Productivity
PLT
Platform Services
Workflow, Notification, and Document Management services deployed as supporting backend capabilities
HPA
AKS Runtime Controls
Private cluster, containerized services, namespace isolation, and horizontal scaling for backend workloads
DATA & STORAGE LAYER
Clear separation between enterprise analytical data, CMS operational data, and file-based artifacts
Microsoft Fabric
Microsoft Fabric
Read-only enterprise analytics source consumed only through backend services using governed backend access
Azure DocumentDB
Azure DocumentDB
CMS operational datastore for workflow state, requests, notifications, document metadata, settings, and audit records
Azure Storage Account (Blob Storage)
Azure Storage Account (Blob Storage)
File and artifact storage for exports, generated documents, attachments, and archived workflow artifacts
IDENTITY, SECURITY & OPERATIONS
Cross-cutting controls applied across independently deployable backend services
Azure AD
Azure AD SSO
Enterprise authentication, token issuance, and identity context for frontend and backend API access
RBAC
Backend Authorization
RBAC + ABAC enforced directly within backend services for protected APIs, workflow operations, and document access
Datadog
Datadog
Centralized logs, metrics, traces, monitoring, alerting, and observability across platform components
Infobip
Infobip
External outbound notification delivery for email and workflow event messaging
DEPLOYMENT PRINCIPLES
1
Separate Deployables
React frontend and AKS backend services are deployed separately to support release isolation and independent scaling
2
No BFF Layer
No dedicated backend-for-frontend component is deployed; secured backend services directly expose the required frontend APIs
3
Private Backend Exposure
AKS remains private and reachable only through protected origin controls behind App Gateway and Akamai
4
Data Separation
Fabric stays read-only for enterprise analytics, DocumentDB stores CMS operational data, and ADLS stores binary artifacts
User access & edge layer Frontend deployment Backend compute on AKS Data & storage layer Identity, security & operations Deployment principles