01 / Project showcase

Project showcase

Platforms and tools I've worked on, from internal developer platforms to upstream Kubernetes projects and smaller open source utilities.

I've included my role on each one because "worked on" can mean almost anything. Some company work cannot be linked publicly, so I've kept those entries to details I can stand behind.

Project showcase

Open source 2024 to present Active

Created · Creator and maintainer

Kubernetes SSH Router

A small Go service that maps SSH users to controlled Kubernetes exec sessions without distributing kubeconfig files.

I built it to explore a Kubernetes-native alternative to another static bastion configuration. User definitions live in Secrets and routing changes are picked up by resource watchers.

It includes Prometheus metrics, horizontally scalable deployment support and unit-testable packages. It remains an active engineering project rather than a finished product.

Work included

  • Kubernetes exec routing with client-go
  • Secret-backed users
  • Resource watchers and Prometheus metrics
  • Scaling support and unit tests
  • Go
  • client-go
  • Kubernetes
  • SSH
  • Prometheus

Project links

Open source testing Current Active

Created · Creator and maintainer

Terratest Utils

Go helpers for Terratest-style E2E tests of Kubernetes custom resources and operators.

The harness covers the repeated parts of operator testing, especially setup and assertions around resources managed by Argo CD, Rollouts, Workflows, cert-manager, Flux and External Secrets.

Work included

  • Reusable Go packages for Kubernetes E2E test suites
  • Helpers for custom resources and operator behaviour
  • Test coverage aimed at real cluster interactions
  • Go
  • Kubernetes
  • Terratest
  • Argo CD
  • cert-manager
  • Flux

Project links

Open source tooling Current Active

Created · Creator and maintainer

Komodor Klaudia Sync

A Go-based sync tool that sends local Klaudia documentation into Komodor from GitHub Actions or another CI job.

I built it so Klaudia.md files can live with the code they document, while a containerised sync step keeps Komodor updated with create, update and delete behaviour.

Work included

  • GitHub Action and Docker image for CI-based sync
  • CRUD sync between local documentation and Komodor
  • Example repository for wiring it into a workflow
  • Go
  • Docker
  • GitHub Actions
  • Komodor
  • Klaudia
Open source tooling Current Active

Created · Creator and maintainer

Komodor Security Reporter

A Go service that connects running Kubernetes workloads, scanner findings from Trivy, Snyk or Wiz, and Komodor.

The aim is low-noise reporting: match scanner output to workloads and send the useful parts into Komodor without turning every scan into more manual triage.

Work included

  • Kubernetes workload discovery
  • Security scanner report ingestion
  • Komodor reporting from matched findings
  • Go
  • Kubernetes
  • Trivy
  • Snyk
  • Wiz
  • Komodor

Project links

Personal tooling Current Active

Created · Creator and maintainer

Zwift iCal

A Go service that publishes upcoming Zwift events as iCal feeds for calendar apps.

It turns Zwift events into subscribable feeds for Google Calendar, Apple Calendar, Outlook and other clients that understand .ics files.

Work included

  • Zwift event collection
  • Auto-updating iCal feed generation
  • Containerised deployment
  • Go
  • Docker
  • iCalendar
  • Zwift
  • Prometheus

Project links

Kubernetes operations Current Active

Created · Creator and maintainer

Stuck Pod Reaper

A Kubernetes cleanup tool for Pending pods stuck in ContainerCreating or PodInitializing beyond a chosen threshold.

It can run in dry-run mode, scan one namespace or all namespaces, and only targets pods with owners so controllers can recreate them.

Work included

  • Pending pod detection through client-go
  • Namespace-scoped or cluster-wide scans
  • Dry-run and delete modes for safer rollout
  • Go
  • Kubernetes
  • client-go
  • Docker

Project links

Kubernetes operations Current Active

Created · Creator and maintainer

System Upgrade OS Patch

A small image for Rancher System Upgrade Controller that gives weekly node patch jobs a changing version endpoint.

The controller can use the generated year, month and week value to trigger regular node upgrades, keeping operating system patches on a predictable cadence.

Work included

  • Weekly version endpoint for upgrade plans
  • Container image for scheduled OS patch workflows
  • Integration point for Rancher System Upgrade Controller
  • Go
  • Docker
  • Kubernetes
  • Rancher
Homelab tooling Current Active

Created · Creator and maintainer

UniFi Cert Updater

A Kubernetes-aware updater that reads a TLS secret and deploys that certificate to a UniFi gateway or controller.

I use it for a Let's Encrypt wildcard certificate from my k8s/k3s cluster and a UniFi Dream Machine Pro. It checks fingerprints, activates the new certificate and trims older certificates.

Work included

  • TLS material loading from Kubernetes Secrets
  • UniFi certificate upload and activation
  • Old certificate cleanup with a configurable limit
  • Go
  • Kubernetes
  • UniFi
  • Let's Encrypt
  • Docker

Project links

Homelab observability Current Active

Created · Creator and maintainer

VM Hub 5 Exporter

A Prometheus exporter for Virgin Media Hub 5 router metrics.

It polls the Hub 5 local API and exposes provisioning and PON status metrics, with a Grafana dashboard kept alongside the container build.

Work included

  • Prometheus metrics endpoint
  • Hub 5 provisioning and PON status polling
  • Grafana dashboard and container image
  • Go
  • Prometheus
  • Grafana
  • Docker
  • Virgin Media Hub 5

Project links

Platform engineering 2017 to 2019 Delivered

Contributed · Infrastructure platform engineer

On-premises Kubernetes Platform

Platform engineering for Kubernetes clusters running across private infrastructure and AWS.

The private clusters needed some of the lifecycle behaviour people expect from public cloud, but had to work with physical failure domains and the organisation's existing provisioning systems. I worked on that gap through availability zones, a custom cloud controller and maintenance automation.

Work included

  • Availability-zone support and a custom cloud controller for node provisioning
  • Automated patching, health checks, reporting and infrastructure deployment
  • Go
  • Kubernetes
  • Terraform
  • Prometheus
  • On-premises infrastructure

03 / Contact

Want to ask about one of these projects?

Email me if you'd like more detail, or if you're working through a similar problem.

Start a conversation