Architecting Cloud Solutions That Last

Challenge

An enterprise Appian deployment with no centralized architecture — backend services exposed directly to the internet, all environments sharing a single subscription, no WAF, and entirely manual operations with no governance framework.

Approach

Re-architected the platform into a structured Azure landing zone with defense-in-depth security (App Gateway + WAF + Azure Firewall), a 3-node HA cluster, RBAC with MFA, CI/CD pipelines, and a full governance framework covering naming, tagging, and access controls.

Challenge

No standardised cloud foundation - each team provisioned ad-hoc subscriptions with inconsistent security controls and no governance layer, making audits and compliance near-impossible.

Approach

Designed a hub-spoke landing zone across 3 Azure subscriptions with centralised networking, management groups, and full Terraform automation backed by Azure Policy at every boundary.

Challenge

A fragmented single-subscription Azure deployment with all environments co-located, services exposed directly to the internet, no WAF or centralized firewall, and no governance — with a hard requirement to migrate workloads to a new subscription architecture without service disruption.

Approach

Designed a multi-subscription Azure landing zone with a Zero Trust security model, executed a structured cross-subscription migration with zero downtime, and embedded Microsoft Defender for Cloud and Azure Sentinel for continuous threat detection and compliance.

Challenge

Three independently deployed security tools — Carbon Black for endpoints, Cisco Umbrella for DNS/network, and Defender for Servers for workloads — generating telemetry in complete isolation. No cross-platform correlation, no unified visibility, and threat investigation requiring manual pivoting between disconnected dashboards.

Approach

Designed and built a centralized SIEM architecture on Azure Sentinel, integrating all three security toolchains via serverless Azure Function pipelines, automated incident response via Logic Apps, and KQL-based analytics rules for correlated multi-source threat detection.

Challenge

Release cycles averaging 5 days with 6 manual approval gates, frequent rollback incidents, and no observability into what was deployed where - causing production outages and developer frustration.

Approach

Built end-to-end Azure DevOps pipelines with containerised builds, AKS deployment, automated security scanning, and Datadog integration - eliminating all manual gates outside of a single production approval.

Challenge

Always-on privileged accounts across a hybrid server estate, no just-in-time access controls, manual and inconsistent server configuration causing security drift, and no automated path to deploy security tooling at scale across Azure, on-premises, and multi-cloud infrastructure.

Approach

Implemented Azure Privileged Identity Management with just-in-time access and approval workflows, automated Defender for Servers deployment across the full hybrid estate using Ansible playbooks, and established a Zero Trust security model where identity is enforced as the primary control boundary.