Socheat
NixOS Consultant and DevOps Expert · UTC+7
I lead an infrastructure and DevOps team and still do much of the hands-on work myself.
I was a senior frontend and backend engineer before that, so I treat a platform as a product for the developers who use it. I build setups that are declarative, reproducible and boring to operate, and document them so your team can run them without me.
Available for consulting
- NixOS
- Servers and fleets defined in one flake: reproducible builds, declarative deployments to many hosts, secrets kept out of the Nix store, atomic upgrades with rollback, and migrating existing Debian or Ubuntu machines to NixOS.
- DevOps
- Getting code from a commit to production safely: CI/CD with GitHub Actions or GitLab CI, container builds, GitOps with Argo CD or Flux, and metrics, logs and alerts with Prometheus, Grafana and Loki so problems show up before users report them.
- Bare Metal Kubernetes
- Kubernetes on your own hardware with k3s, RKE2 or Harvester, on Proxmox or bare metal: cluster design and bootstrap, networking and load balancing without a cloud provider, persistent storage, and day-2 operations such as upgrades, backups and monitoring.