SiteTidy
A practical web quality toolkit for audits, favicons, sitemaps, robots and the sort of boring jobs developers would rather automate.
Actually shipped. Miracles happen.Open project ↗Hi. I’ve spent more than 20 years turning awkward business problems, half-formed ideas and “couldn’t we just…” conversations into software that actually works.
SELECTED MISSIONS
Released work gets a name and a link. The things still in the workshop stay deliberately vague until they’re ready to escape.
A practical web quality toolkit for audits, favicons, sitemaps, robots and the sort of boring jobs developers would rather automate.
Actually shipped. Miracles happen.Open project ↗Exploring how AI agents, applications and the web can interact through cleaner protocols and narrower, more intentional interfaces.
Reasonably ambitious. Possibly unhinged.Still in the workshopExperimenting with agentic automation, repeatable computer work and orchestrated runtimes that can do useful things without constant babysitting.
Because apparently macros weren’t enough.Still in the workshopBuilding with small local models and private, on-device intelligence to see where AI is more useful when the cloud is not invited.
Small models. Surprisingly big rabbit holes.Still in the workshopA collection of experiments around search, local data and everyday questions that should be much easier for the internet to answer.
The web still has plenty of stupid little problems.Still in the workshopSmall tools, widgets and utilities designed to answer a question or solve a job without making the user fight through a content farm first.
Useful beats impressive.Still in the workshopCHARACTER SHEET
I started out building PHP-based web applications and CMS platforms. It was my way into professional software development rather than the technology that came to define my career.
I moved into the Microsoft/.NET world relatively early and have spent most of my career there, building everything from business applications and internal platforms to large, interconnected systems that businesses depend on every day. Over the years that expanded well beyond writing software into architecture, infrastructure, networks, security, technical leadership and, increasingly, product development.
At WageDayAdvance I became IT Director while the business was scaling at serious speed. The company reached No. 8 in The Sunday Times Tech Track 100 ↗ — the ranking of Britain’s fastest-growing private technology companies. At that pace, technology couldn’t simply support the business. It had to anticipate where the business was going and be ready before it got there.
I’ve led large technical teams, but I’ve always preferred staying close to the technology. I’ve designed and built bespoke in-house platforms, operational systems, integrations and much of the less glamorous machinery that businesses quietly depend on — the things nobody talks about until they stop working and suddenly become the most important system in the building.
That has included being hands-on with virtualised clusters, servers, networks and major infrastructure projects as well as software architecture. I’ve never particularly liked the idea that becoming an architect means eventually graduating from actually building things. A diagram is useful. Knowing whether the thing on the diagram will actually work is rather more useful.
I’ve also headed responses to ransomware and other serious security incidents. Those situations have a habit of stripping technology back to fundamentals very quickly: understand what’s actually happening, contain it, make sensible decisions with incomplete information, protect what still works and get the business safely moving again.
Today I’m back where I’m happiest: building things. My own stack has settled around .NET/C#, Vue.js and modern web technologies, with whatever databases, infrastructure, cloud services and AI tooling make sense for the problem. I’m particularly interested in the point where software engineering, AI and automation meet — not AI for the sake of putting an AI button on something, but using it where it genuinely changes what software can do.
After more than two decades in technology, I’ve been a developer, architect, infrastructure engineer, technical leader and director. These days the title matters considerably less to me than the problem.
I’m an entrepreneur by nature, but probably a problem solver first. Give me something difficult, messy or apparently impossible and my instinct is still the same as when I first started writing software: take it apart, understand how it works, and figure out a better way to build it.
Not a toy MVP. Not six months of diagrams. Something real enough that a user can trust it.
If SQLite solves it, use SQLite. Nobody gets bonus points because the architecture diagram required A3 paper.
Strategy is useful. So is being able to open the logs, trace the traffic and understand what the machines are actually doing.
LOADOUT
Tools change. Good judgement is slower to deprecate.
THE DAY JOB / THE COMPANY
I build products and solve technical problems through SevenVectors. If you’ve got something ambitious, awkward or technically interesting, that’s the front door.
TRANSMISSIONS
The most dangerous thing humanity has built may not be artificial intelligence itself. It may be the race that makes slowing it down almost impossible.
AI taking over may not look like a robot uprising at all. It may simply be what happens when a faster, more capable form of intelligence inherits systems built by a slower one.
Modern AI can explore software with something closer to human persistence. Our defensive assumptions need to catch up.
OPEN CHANNEL
Good. Those are usually the interesting ones.