Skip to content

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/ and docs/release2/evidence/A2-awx-control-plane/.
  • Private platform services - kubernetes/, docs/release2/evidence/O4/, and docs/release2/evidence/O5/.
  • Monitoring and resilience - P7/, P8/, P9a/, P9b/, P9b-redesign/, and BCDR documentation.
  • AI operations - docs/release2/evidence/O6/ and companion local-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.

Release 3 Roadmap