The person who gets alongside Devs, Ops, InfoSec, and Marketing — and builds the automated tooling that makes everyone's lives easier.
Developers are focused on shipping features and hitting sprint goals. They need help making that process smooth and safe.
Ops are watching the AWS bill, worrying about maintainability, hoping they won't get woken up at 3am.
InfoSec & Governance are responsible for security and data integrity — but often don't have many levers to pull.
Marketing just want the campaign out the door, but all the red tape makes it feel impossible.
The rest of the business needs to be listened to — and where there are efficiencies to be found... so who helps?
DevOps — or DevSecOps, Platform Engineering, or "Sys Admin who has hung around with Devs too long" — gets alongside all these teams and builds automated tooling to make everyone's lives easier.
That's what I do. I've been doing it for over 25 years.
Primary DevOps engineer responsible for transforming the entire technology platform — from small, unmanaged golden image servers to a full Kubernetes platform on AWS. Mentored a junior engineer into a capable DevOps professional. Now leading technical AI adoption across the organisation alongside continued platform optimisation.
Joined Semantico as they were acquired by HighWire Press, then MPS Limited. Grew the DevOps team and learnt a huge amount — containerised ~100 Java and Go applications with Docker Swarm, Jenkins and Ansible. Introduced centralised logging, OpenTracing and security scanning. Stepped up to manage the UK Operations team across Belfast, Brighton and Los Gatos.
Started as a Junior Sys Admin at Pinnacle, one of the UK's first B2B ISPs — installed their first datacentre server and was promoted within six months. When Pinnacle was acquired by Precedent in 2003, grew the hosting offering into a profitable multi-datacentre, geographically disparate service with offices across the UK and Australia. Customers included BMA, Parkinson's UK, Wildlife Trust, British Heart Foundation and Anthony Nolan.
Where it all started. Also: Computing, Maths & Physics A-levels at Worthing Sixth Form (1994–1996).
| BSc Computer Science | University of Kent, 1997–2000 |
| A Levels (Computing, Maths, Physics) | Worthing Sixth Form, 1994–96 |
| Leading With Confidence | Bluecrest Wellness |
| Bluecrest Academy of Management | Bluecrest Wellness |
| AWS Business Professional | Amazon Web Services |
| OWASP Top 10 | LinkedIn Learning |
Golden image servers → Docker Swarm → Kubernetes on AWS EKS. Designed auto-scaling with Karpenter, multi-environment deployments, and Helm-based release management. Each step was deliberate — proving value before moving to the next level of complexity.
Replaced a fragile deployment process with proper CI/CD pipelines for every service — from PHP APIs to React frontends. Developers went from manual deployments to deploying via pull request with automated testing, linting and security scanning.
Introduced full observability where almost none existed — Prometheus, Grafana, Loki for logs, NewRelic for APM, uptime monitoring and alerting. Teams can now see what's happening in real time rather than finding out from customers.
Transformed the security posture from unmanaged to achieving strong compliance against ISO 27001:2022. Significantly improved AWS CIS benchmark scores. Introduced vulnerability scanning, dependency auditing and security-by-default practices.
All infrastructure managed via Terraform and Ansible — from Kubernetes clusters to DNS records. Reproducible, auditable, and version-controlled. No more snowflake servers.
Empowered developers to own their deployments. Introduced environment parity, feature branching, contract testing and a hackathon platform for innovation. Built a culture where "it works on my machine" is no longer acceptable.
Built a multi-agent pipeline that autonomously audits an entire software estate — investigating every git repo, identifying software versions, checking for end-of-life risks, and producing a prioritised upgrade roadmap. Agents self-improve between passes.
Created tooling that uses AI to build ISO 27001 evidence matrices — mapping controls to real evidence, generating audit preparation materials, and identifying gaps. Turned months of manual compliance work into days.
Developed a structured "AI-first with human verification" framework for DevOps troubleshooting — with reusable skills for querying Grafana, Kubernetes, AWS, and Jira. AI does the heavy lifting; humans make the decisions.
Part of the team that authored the organisation's AI policy. Personally built the reusable skills library — enabling consistent, safe AI-assisted development across multiple teams. Skills cover everything from ticket creation to TDD to infrastructure debugging.
Gave developers a one-command local Kubernetes environment. Instead of a 40-step README that nobody followed, they type ./up.sh and get a full stack with hot-reload. The number of "it works on my machine" conversations dropped to near zero.
Replaced the Kubernetes Cluster Autoscaler with Karpenter and watched nodes spin up in seconds instead of minutes. The right-sizing alone justified the migration — no more paying for nodes that are 20% utilised because the autoscaler can't scale down gracefully.
For a small team, the operational overhead of an Elasticsearch cluster is brutal. Loki gives us grep-at-scale with a fraction of the resource cost — and it speaks the same language as our Grafana dashboards.
Templated 20+ microservices without losing sanity. A shared base chart with per-service values means a new service goes from "code complete" to deployed in all environments in under an hour. The trick is knowing when not to abstract.
The first step wasn't Kubernetes — it was convincing everyone that "we'll just update the server" isn't a deployment strategy. Docker gave us reproducibility. Once people saw that a container built locally was identical to production, the rest followed naturally.
When you're the primary DevOps engineer, you can't afford to spend three days researching every decision. AI assistants became my rubber duck, my second opinion, and my documentation writer. The key is knowing what to trust and what to verify.
Fantastic family. Gaming. (Watching) football. Podcasts. Many things computer-based.
I've helped lead Breakfast Church (in various forms) at Maybridge with a great team of people for over 26 years.
I live in Worthing, West Sussex — close enough to the sea to feel smug about it, far enough from London to avoid the commute.
Started doing extra walks for a workplace step challenge for Diabetes UK — and discovered that if you find an unputdownable audiobook, you have to keep walking to find out what happens next.
I now average a couple of books a month along Worthing beach. This year's themes: epic fantasy, comedians with autism, and dystopian comedies.
| 🗡️ | Brandon Sanderson — Mistborn trilogy, started Stormlight |
| 🗡️ | Terry Pratchett — 4 Guards rereads this year (via Spotify) |
| 😂 | Fern Brady — Strong Female Character |
| 😂 | Pierre Novellie — Why Can't I Just Enjoy Things? |
| 😂 | Richard Ayoade — Ayoade on Top |
| 🌍 | Jasper Fforde — Shades of Grey, Red Side Story, Early Riser |
| ✝️ | Brian McLaren — We Make the Road by Walking |
| ✝️ | Rob Bell — Love Wins |
| 🪦 | Hilary Mantel — Wolf Hall (should have done History at GCSE) |
"There are only two hard things in Computer Science: cache invalidation, naming things, and off-by-one errors."
— Phil Karlton (with anonymous extension)"We replaced our monolith with micro services so that every outage could be more like a murder mystery."
— @HonestUpdate"Make it work, make it right, make it fast."
— Kent Beck"Fast, Cheap, Good — pick two."
— The Project Management TriangleInterested in working together, or just want to chat about DevOps, AI, or Kubernetes on everything? Drop me a message.