- Infrastructure written down as codeTerraform · AWS
- Built once, shipped exactly as testedActions · Docker · GHCR
- Git holds the truth; the cluster followsFlux CD · Kustomize
- Restarts what fails, on its ownKubernetes · k3s
- Secrets encrypted, never in a text fileSOPS · age
- Problems seen before anyone reports themPrometheus · Grafana
- A bad release rolls back with a revertFlux CD · Git
- Left so the team can run it without meHandover
How I work
Listen first.
The work starts with understanding what needs fixing, from the people who live with the system every day.
Use what’s there.
Budgets are real and people wear several hats. I work with what’s already available and add complexity only when it solves a real problem.
Leave it running.
The best setup is one the team can keep running, long after I’ve stepped away.
Every tool, one job
Terraform writes the foundations down on AWS. GitHub Actions builds each release once, Docker seals it and GHCR keeps it. Kubernetes runs it, on lean k3s clusters that Flux keeps matching Git. Kustomize keeps environments honest and SOPS keeps secrets secret. Prometheus, Loki and Grafana make sure the team hears first. Underneath it all, Node.js services I write myself.
The path
From IT support to platform engineering, every break mended.
- 2015BCA, Computer ScienceSpicer Adventist University
- 2019IT support & web developerTier 1 and 2 support, websites, internal tools
- 2020MBA, Information TechnologyAdventist International Institute of Advanced Studies
- 2023IT administrator & web leadNetworks, devices, access, web projects
- 2025DevOps & backend engineerAdventist World Radio · contract, remote