About

About me

I am a Site Reliability Engineer. I started in operations in 2020, looking after Linux hosts across several cloud providers, and moved from there into Kubernetes platforms, infrastructure as code and observability on AWS and GCP.

Two habits shape how I work: I measure before I change anything, and I move manual operations into Git and pipelines so that the next change is reviewed and repeatable.

Background

Six years in operations and reliability across AWS and GCP, from a fleet of Linux VMs to Kubernetes platforms and GPU infrastructure for ML teams.

Focus areas

  • Reliability and incident response. Rollouts that do not drop requests, incidents followed through to root cause, and every mitigation committed to Git.
  • Kubernetes platforms. EKS and GKE: version upgrades, autoscaling, disruption budgets, network policy and admission guardrails.
  • Infrastructure as code and delivery. Terraform and OpenTofu with a plan on every merge request, drift detection, and GitOps delivery through Argo CD.
  • Observability. Metrics, logs and traces that connect to each other, plus monitoring for the monitoring stack itself.

Tech stack

  • Kubernetes and containers: EKS, GKE, Helm, Cluster Autoscaler, HPA / PDB, NetworkPolicy, External Secrets, Docker
  • IaC and automation: Terraform, OpenTofu, Ansible, Python, Shell, Go
  • CI/CD and GitOps: Argo CD, GitLab CI, GitHub Actions, Jenkins
  • Observability: Prometheus, Grafana, Loki, Tempo, OpenTelemetry, ELK
  • Networking and edge: ALB / Ingress, Nginx, Istio, DNS, CloudFront, Cloudflare
  • Databases and streaming: PostgreSQL, MySQL, Redis, OpenSearch, MongoDB, Kafka
  • Cloud and AI infrastructure: AWS, GCP, GPU nodes (H100), GPU time-slicing

Education

Shih Chien University, B.S. in Information Technology and Communication, 2015 – 2019.

Why this site exists

Commit history records what changed. It does not record why. I write here so that the reasoning behind a decision is still available after the details have faded.

Contact

This site

Built with Astro as a static site. Posts are Markdown files in a content collection.