Hi, I’m Rob. I build the quiet infrastructure the rest of an engineering organisation stands on — the paved roads, golden paths, and shared platforms that let teams move faster without having to think much about the ground beneath them.
These days most of my time goes to AI-platform engineering: the operations layer underneath AI-assisted development. The platforms underneath the platforms.
How I got here#
I’ve been curious about computers since I was a kid, and it started with games. One of my friends was always a few years ahead of the rest of us — in the 90s he’d wired his and his brother’s machines together over a LAN so we could all play, then burned me CDs of games, then ran a Team Fortress server on a laptop humming away beside his main PC. I had no idea how any of it worked, and I desperately wanted to. It just seemed like the coolest thing in the world.
I had no idea how any of it worked, and I desperately wanted to.
I did my IT apprenticeship in London and took my first job in Dorking, fixing clients’ computers and Windows servers. From there I looked after a recruitment agency’s machines — and met a developer for the first time. He showed me what you could do with a scripting language, PowerShell in my case, and something clicked: all the boring, repetitive work I did by hand could be written down once and handed to the computer. I never saw the job the same way again.
A software house near London Bridge taught me virtualisation, the software development lifecycle, and my first real taste of AWS. I wrote custom PowerShell modules and bash scripts, and built a ghost-imaging server so setting up a new starter’s machine stopped eating an afternoon. At Inspired Thinking Group I moved into cloud operations across GCP and Azure, picked up Python, and built a Flask app to manage our Terraform deployments — an internal developer platform before I knew that was the name for it.
My real trial by fire was re-architecting our cloud-platform monorepo so its Terraform state could deploy from scratch with no circular dependencies. It took a year, and it taught me the worth of design sessions and Architecture Decision Records more than any book could. By the end the platform was scalable and maintainable — and no developer had to touch Terraform at all. Everything ran through config YAML that behaved the way Helm values do: sensible defaults, overridable per environment, DRY throughout. I put the same care into the pipelines around it — Terratest, Checkov scanning, linting, schema validation, every module gated on tests and a proper README before it could reach main. Somewhere in there I fell for the GitOps way of working with ArgoCD and Kubernetes, and got to architect our monitoring stack on Grafana, Jaeger, Prometheus and Thanos.
Since then the work has been less about any single platform and more about how a whole organisation builds: a CI/CD job catalogue for seventeen-odd teams to share, golden paths and Helm templates, and most recently a Claude marketplace of shared skills anyone in the company can use — the same DRY instincts, sensible defaults, customisable per repo or per person. Just as much of it now is spent with people: a DevOps champions group, an AI champions workgroup, and knowledge-sharing sessions and mentoring to help teams move faster while staying secure.
What I care about#
I care about golden paths and guardrails: giving infrastructure enough structure to stay aligned without ever getting in a development team’s way. The best version of that is quietly customisable — which is why, in the end, I care most about feedback. A platform is only as good as the teams who actually want to use it, and they’ll tell you what’s working if you build the room for them to. The best of it lifts a whole team without ever calling attention to itself — and this site is where I write down what I learn trying to get there.
On AI#
AI is, genuinely, the thing I dreamed about as a child. Talking to a computer in plain language is straight out of the cartoons I grew up on, and I don’t think the wonder of it will ever quite wear off. More than that, I think it lowers the bar to innovation — handing more people the means to solve real, human problems in clever and imaginative ways. Building the operations layer underneath all that, so it becomes something others can actually trust, is the most interesting work I’ve done.
Away from the keyboard#
I’m based in Herefordshire, and a good deal of my spare time goes to the countryside around me — travelling, gardening, and hiking in the hills; mountains always give often-needed perspective. The rest of it usually finds me with a book in my hands, or a few moves from losing a game of chess.
What you’ll find here#
- Posts — field notes on platform engineering, AI-platform engineering, infrastructure-as-code, and the patterns that hold (or don’t) at scale.
- Case Studies — the longer arcs: a problem in production, the work it took to fix, and what changed once it shipped.
- Tools — the systems I use daily and recommend.