The boring, reliable plumbing under modern software.
I help teams move from “it works on my laptop” to a system that deploys, scales, recovers and alerts by itself. Containers, clusters, CI/CD, observability and the cloud plumbing to glue them together — designed to be handed over, not locked in.
Six things I routinely own end-to-end for clients.
Container & Kubernetes
Dockerised workloads, Helm charts, cluster design, RBAC, autoscaling, and zero-downtime rollouts.
Cloud migration
AWS, GCP, Hetzner and on-prem — right-sized, cost-aware moves with sensible exit strategies.
Infrastructure as code
Terraform, Ansible and Pulumi. Reproducible environments, code-reviewed infrastructure.
CI/CD pipelines
GitLab CI, GitHub Actions and self-hosted runners. Fast, traceable deploys you can trust.
Observability & on-call
Prometheus, Grafana, Loki, Sentry. SLOs that mean something and alerts that rarely lie.
Backup & disaster recovery
Tested, automated backups. Restore drills. Runbooks written in plain English, not tribal memory.
A predictable, low-drama engagement.
-
01
Audit
A week of reading code, diagrams, dashboards and paging history. Output: a written assessment with prioritised risks.
-
02
Plan
We agree on scope and a sequence of reversible steps, each with a rollback.
-
03
Ship
Changes land in small, reviewable PRs. Nothing I do is a surprise on Monday morning.
-
04
Hand-over
Runbooks, architecture diagrams, and a pairing session so your team owns what I built.