Tarmizi Hairudin A.

I keep production systems running and I can show you numbers, not adjectives. A recent example: a production Kubernetes cluster where every system component had quietly stopped scheduling. Cluster-wide DNS was dead and the usual debugging tools were among the first casualties. I found the cause, restored service in about two minutes, then fixed the layer that actually caused it and left behind monitoring that cut detection time from 14 hours to 5 minutes. BACKGROUND 9 years in IT, 7 of them in DevOps and cloud infrastructure including 5 years as Infrastructure Lead, responsible for production architecture rather than executing someone else's design. Currently completing a Master's in Information Systems, with research on CI/CD implementation published in a peer-reviewed journal. WHAT I DO - Kubernetes : production clusters on GKE and self-managed (kubeadm, Rancher). Autoscaling, persistent storage, incident response, cost optimisation, workload placement. - AWS : ECS Fargate behind ALB, RDS, S3 with CloudFront, IoT Core with device shadows, X-Ray, CloudWatch, IAM/VPC/WAF/Secrets Manager. Also Google Cloud and extensive Alibaba Cloud. - Infrastructure as Code : Terraform and Ansible. Infrastructure defined in version control rather than assembled by hand, so changes are reviewable and environments can be rebuilt instead of repaired. - CI/CD : Jenkins, GitLab CI, GitHub Actions, AWS CodeBuild, GCP Cloud Build, ArgoCD. On one engagement: deployment time 2–3 hours → 15 minutes, success rate 70% → 95%, rollbacks 25% → 5%. Measured during rollout and published in a peer-reviewed journal, not self-reported. - Disaster recovery : DC/DRC architecture with failover actually tested, not just documented. - Security & compliance : ISO 27001 controls, WAF, least-privilege IAM, SonarQube in the pipeline, hardened Linux. - Monitoring : Grafana, Prometheus, ELK, Zabbix, CloudWatch. SELECTED WORK Built an entire AWS environment for an IoT product company, from empty account to handover: containerised their APIs onto ECS Fargate behind a load balancer, migrated object storage and message queuing to managed services, configured IoT Core with device shadows for per-device settings, built the CI/CD pipelines, ran a security audit, tuned cost, and finished with diagrams, runbooks and knowledge transfer to their own team. Infrastructure Lead for five years at a Jakarta company: designed and ran the primary and disaster-recovery data centres with tested automatic failover, moved workloads to Kubernetes under Rancher, and implemented ISO 27001 controls. HOW I WORK I find out why your current setup breaks before I automate it. Automating a broken process just produces failures faster. I tell you what I'm trading away and why. Cheap preemptible infrastructure is right for some workloads and reckless for others. I'll tell you which of yours is which, and I'll withdraw my own suggestion if the numbers don't support it. I'd rather the fix hold than look impressive. On that cluster outage, the first repair worked and then the fault returned, because the setting lived one layer deeper than where I'd corrected it. I now treat any hand-applied change as temporary until it exists in configuration which is most of why I work in Terraform and Ansible rather than on the server. I finish properly. Documentation, runbooks and handover are part of the job, not an optional extra you should be able to run what I build without me. If you have something broken, something slow, or something you're afraid to deploy on a Friday afternoon tell me what's happening and I'll tell you honestly whether I'm the right person for it. On Upwork: 22 projects delivered, 100% Job Success.