AMportfolio@v3.0.0
๐Ÿ‘‹ Ask me anything!
Back to Home

Case Studies.

Real Azure engineering work โ€” infrastructure, observability, and security architecture built in production.

AKS ยท CI/CD ยท DevOps

AKS Deployment with Azure DevOps CI/CD

Built a complete containerized application lifecycle on Azure โ€” from Docker build to ACR image storage, AKS deployment, and full production observability with Prometheus and Grafana. The Azure DevOps pipeline automates build, push, and deploy on every commit. Kubernetes Ingress handles external routing. AlertManager is configured for CPU, memory, pod crash, and replica mismatch alerts.

โœ“2 replicas running โ€” zero manual deploys
โœ“Prometheus + Grafana โ€” full cluster observability
โœ“Automated pipeline โ€” commit to live in under 3 mins
Challenge Solved

Resolved exec format error (CrashLoopBackOff) caused by a Windows-built Docker image running on Linux AKS nodes โ€” fixed by switching the pipeline build agent to ubuntu-latest.

Azure DevOpsDockerAKSACRKubernetesPrometheusGrafanaHelmNGINX IngressYAML
RG: rg-aks-devops
Region: East US
AKS Cluster: aks-platform
ACR: acrplatformpro3.azurecr.io
Read Full Case Study
Pipeline Stages
1
Checkout Code
Pull latest from repo
2
Build Docker Image
ubuntu-latest agent
3
Push to ACR
Tagged BuildId + latest
4
Deploy to AKS
kubectl apply -f
5
Rollout Status
kubectl rollout status
6
Live โœ“
NGINX Ingress โ†’ Internet
Azure Monitor ยท Cost Management ยท Automation

Azure Monitoring & Cost Optimization Platform

Designed and deployed a centralized Azure monitoring platform across a full resource estate. Unified logging via Log Analytics Workspace, real-time alerting via Azure Monitor, and cost governance via Azure Cost Management and Advisor. Automation Runbooks handle auto-remediation โ€” when CPU threshold breaches, the runbook executes and restarts the service automatically.

โœ“70% reduction in manual ops via IaC and Runbooks
โœ“30% cloud cost savings identified via Advisor
โœ“Budget alerts at 50% / 75% / 90% thresholds
Challenge Solved

Tamed alert fatigue by implementing threshold tuning and log retention policies โ€” reduced noise while maintaining full coverage across all resource types.

Azure MonitorLog AnalyticsApplication InsightsCost ManagementAzure AdvisorAutomation RunbooksKQLAction GroupsGrafana
RG: rg-monitoring-lab
Region: Canada Central
Log Analytics WS: law-monitoring-lab
App Insights: appi-monitoring-lab
Read Full Case Study
Architecture ยท Networking ยท Security

Azure Three-Tier Application Architecture

Designed and deployed a production-style three-tier web application on Azure โ€” React frontend, Node.js/Express backend, Azure SQL database. All inbound traffic routes through Azure Application Gateway with path-based routing. Network Security Groups enforce least-privilege access. No direct internet exposure to backend or database at any point.

โœ“Zero direct DB exposure โ€” backend-only SQL access
โœ“Auto-scaling: 2โ€“10 instances, CPU threshold 70%
โœ“2/2 backend health probes returning 200 OK
Challenge Solved

Fixed backend health probe failures by implementing a dedicated /api/health endpoint returning 200 OK โ€” resolved the missing health check that caused App Gateway to mark the backend pool as unhealthy.

ReactNode.jsExpressAzure SQLApplication Gateway v2Azure App ServiceNSGApplication InsightsRBACVNet
RG: rg-multitier-autoscale-dev
Region: Mexico Central
App Gateway: agw-multitier-dev (Standard V2)
Zones: 1 / 2 / 3
Read Full Case Study