Release 2: Platform Engineering and Multi-Cloud¶
Status: Implemented and evidenced
Release 2 is implemented and routed to public-safe evidence through Terraform source, GitHub Actions workflow records, evidence folders, screenshots, CLI output, Kubernetes manifests, Ansible/AWX material, architecture documentation, and the companion local AI lab repository.
Release 2 extends the realistic Microsoft hybrid enterprise environment into platform engineering, secure networking, automation, private platform services, operations, and AI governance.
Architecture overview¶
flowchart TD
subgraph Governance["Governance and CI/CD"]
TF["Terraform<br>Multiple isolated roots"]
OIDC["GitHub Actions<br>OIDC and reviewable delivery"]
Trace["Code traceability<br>Source, workflow, proof"]
end
subgraph Network["Network"]
Hub["Hub VNet<br>Azure Firewall<br>FortiGate NVA"]
Spoke["Spoke VNets<br>Private access paths"]
VPN["IPSec and BGP<br>AWS branch routing"]
end
subgraph Platform["Platform Services"]
AKS["Private AKS<br>Kubernetes policy context"]
AVD["Azure Virtual Desktop<br>FSLogix and compliance context"]
Backup["Recovery Services Vault<br>Backup and BCDR"]
end
subgraph Operations["Automation and AI"]
AWX["Ansible and AWX<br>Governed runbooks"]
AIOps["AI Operations Enclave<br>O6 evidence"]
end
OIDC --> TF
TF --> Hub
Trace --> OIDC
Hub --> Spoke
Hub --> VPN
Spoke --> AKS
Spoke --> AVD
AWX --> AKS
AWX --> AVD
Backup --> AVD
AIOps --> AWX
Delivered capabilities - Delivery Engineering¶
-
Multi-root Terraform with isolated state - separate roots divide networking, management, shared services, AKS, AVD, governance, workload, and AWS branch responsibilities.
Terraform State Boundaries -
Secret-less CI/CD - GitHub Actions uses OpenID Connect to support workflow-controlled Azure authentication without routine long-lived deployment credentials.
GitHub Actions OIDC -
Code traceability - source files, workflow evidence, documentation, and proof routes connect platform claims to reviewable implementation evidence.
Code Traceability -
Automation control plane - Ansible and AWX provide governed inventories, runbooks, job templates, controlled execution, and job evidence.
Automation Control Plane
Delivered capabilities - Network Engineering¶
-
Hub-spoke network design - Azure hub-spoke routing, route control, Azure Firewall, and private access paths shape the platform network.
Hybrid Multi-Cloud Networking -
Secure transmission and inspection - FortiGate NVA inspection, Azure Firewall context, route validation, and traffic-path evidence prove more than diagram-level design.
Secure Transmission and Inspection -
IPSec and BGP multi-cloud transit - on-premises routing, Azure VPN Gateway, AWS branch routing, Cisco CSR context, and BGP validation demonstrate dynamic transit.
Hybrid BGP Multi-Cloud Transit
Delivered capabilities - Platform Services¶
-
Private AKS - private platform pattern with Kubernetes manifests, controlled access, network policy context, and operational evidence.
Private AKS Platform -
AVD secure workspace - Azure Virtual Desktop with FSLogix, private access patterns, compliance context, and operator toolchain evidence.
AVD Secure Workspace -
AKS and AVD integration - private platform services are connected through controlled routes and inspected access paths.
Private AKS and AVD Architecture
Delivered capabilities - Operations Engineering¶
-
Monitoring and alert validation - Azure Monitor, Sentinel, Defender for Cloud, and alert evidence demonstrate operational visibility.
Monitoring, Backup and Resilience -
Backup and resilience - Recovery Services Vault controls, backup validation, soft-delete handling, and BCDR planning show recoverability beyond deployment.
Monitoring, Backup and Resilience -
Day-2 automation - Ansible/AWX job evidence and source-controlled runbooks show repeatable operational execution.
Automation Control Plane
Delivered capabilities - AI Governance¶
-
AI Operations Enclave - O6 demonstrates policy-mediated tool use, deny-by-default policy decisions, scoped agent access, network isolation, namespace lifecycle validation, decision logs, and cleanup checks.
AI Operations Enclave -
Companion local AI lab - local reference implementation aligned with O6, using validation, evidence bundles, human review boundaries, and no-auto-apply posture.
Companion Local AI Lab
Capability matrix¶
| Capability | Implementation | Evidence route |
|---|---|---|
| Terraform state boundaries | Multiple Terraform roots, separate ownership boundaries, remote state discipline | terraform/ |
| GitHub Actions OIDC | Workflow-controlled Azure authentication without routine long-lived deployment credentials | .github/workflows/ and OIDC evidence |
| Code traceability | Source, workflow, documentation, and proof routes connected to platform claims | Code Traceability |
| Automation control plane | Ansible, AWX, inventories, job templates, runbooks, and execution evidence | A2-awx-control-plane, ansible/ |
| Hub-spoke networking | Azure hub-spoke routing, Azure Firewall, route control, and private access paths | Release 2 network evidence |
| FortiGate NVA inspection | Inspection path, firewall evidence, and route validation | O1 evidence |
| IPSec and BGP transit | VPN, BGP, AWS branch routing, and cross-cloud validation | P5-vpn, O3b |
| Private AKS | Private platform pattern, Kubernetes manifests, controlled access, and network policy context | O4 evidence, kubernetes/ |
| AVD secure workspace | FSLogix, private access patterns, compliance context, and platform toolchain evidence | O5 evidence |
| AKS and AVD integration | Controlled private platform connectivity and inspected access paths | O4/O5 integration evidence |
| Monitoring and alerting | Azure Monitor, Sentinel, Defender for Cloud, and alert evidence | P7, P8, P9a |
| Backup and BCDR | Recovery Services Vault controls, backup validation, soft-delete handling, and BCDR planning | P9b, P9b-redesign |
| AI Operations Enclave | O6 policy-mediated tool use, scoped agent access, decision logs, namespace lifecycle, and cleanup checks | O6 evidence |
| Companion local AI lab | Local AI infrastructure workflow with validation, evidence bundles, and human review boundaries | local-ai-lab-infra |
Evidence hub¶
Release 2 evidence is organised across multiple repository folders:
- Terraform and delivery -
terraform/,.github/workflows/, OIDC evidence, state-boundary documentation, and code traceability routes. - Network inspection and transit -
docs/release2/evidence/O1/,P5-vpn/,O3b/, and related network validation evidence. - AWX automation -
ansible/anddocs/release2/evidence/A2-awx-control-plane/. - Private platform services -
kubernetes/,docs/release2/evidence/O4/, anddocs/release2/evidence/O5/. - Monitoring and resilience -
P7/,P8/,P9a/,P9b/,P9b-redesign/, and BCDR documentation. - AI operations -
docs/release2/evidence/O6/and companionlocal-ai-lab-infra.
Why it matters¶
Release 2 extends the Microsoft hybrid operating model into a governed platform engineering environment. The engineering signal is the connected lifecycle: delivery, network, private platform services, operations, and AI governance are documented together rather than as isolated services.
Skills demonstrated¶
- Azure platform architecture and governance.
- Infrastructure as Code with multi-root Terraform and isolated state.
- OIDC-based delivery with GitHub Actions.
- Code traceability across source, workflow, documentation, and evidence.
- Hybrid and multi-cloud networking with IPSec, BGP, NVA inspection, and AWS branch routing.
- Private Kubernetes and virtual desktop delivery.
- Operational automation with Ansible and AWX.
- Security monitoring, backup resilience, and BCDR planning.
- Policy-mediated tool use and human approval boundaries in the AI Operations Enclave.
Next step¶
Release 3 carries the platform toward the multi-cloud Kubernetes, GitOps, and DevSecOps roadmap.