The teardown
Ninety minutes. I pull your idea, or the app you've built, apart while you watch. You leave with a written verdict — including the one where I say don't build it.
What I do
I build apps. I also teach people to build their own.
Which is a worse business than only doing the first. I do it anyway, and the rest of this is why.
Keep goingBefore you spend anything
I'll tell you if you shouldn't build it.
Before you've paid me to build it, rather than after. It costs me the project and earns me every referral you make for the next decade.
The part nobody advertises
I'll teach you to do it without me.
Claude Code changed what someone non-technical can ship. Most of this industry is quietly not mentioning that, because billing for the difficulty is still very good money.
Why I'm not a specialist
I'm a generalist, on purpose.
Broad coverage of what's actually moving, instead of depth in one lane. The specialists I know ring me to ask what's happening outside theirs.
How I got that wayClose
GeoCities as a teenager. Years as the office spreadsheet person nobody took seriously. Bubble in 2016, because I could never learn to code the normal way — I needed to see the thing I was building. Then Wappler, Backendless, Webflow. Then Noodl. Then Claude Code.
Each of those was a year or two of being wrong about something. What's left is knowing where every tool breaks, and that's the part you're actually hiring.
On hiring the big firm
The safe pair of hands is usually just the expensive pair of hands.
Think about the last government website you tried to use. Someone was paid extremely well to build that.
The comparison I keep makingClose
A contractor your friend recommends opens the hatch to the attic the big outfit just finished, looks at the insulation, and says: what on earth did they do to you. Then he fixes it for a fraction of the invoice, because doing it right is the part he enjoys.
Big firms are optimised to win contracts, not to make software a person can use — and once yours is signed, you get whoever was free that month. I'm one person. If you genuinely need forty, I'll say so on the first call.
Where this comes from
Ten years teaching adults. About fifty failures.
A decade as a professional trainer, with a Master's-level qualification in designing the training. And a long, well-documented history of getting it wrong.
The failures, specificallyClose
For years I was the spreadsheet guy with no credentials — an emergency solution until a true professional could be found. Then two years on an eLearning platform I abandoned, because I didn't yet know what I didn't know. Then a development company I sank, on good work I'd priced badly. For a long time it felt like building a house with a pile of bricks and good intentions.
That's not here for sympathy. It's why I spot early when a project is heading somewhere bad. The whole route, written up.
What it costs
€90
An hour. Consulting, training, or building — same rate, same brain. I'm quick, which is the part that actually matters.
None of this is charity, by the way. I intend to do very well out of being straight with people.
Four ways in
Most people start with the first.
Claude Code, properly
Live training on your real project. The workflow that holds up past week two, and how to spot the AI being confidently wrong.
Launch readiness
A staged check of what quietly ends apps: untested backups, auth you don't understand, no analytics, no error tracking. Verified means I check it.
Build it with you
I do the parts you can't yet, with you watching, handing each over as you're ready. Plus the call the week after launch.
Things I've built
Go and click on them. Two are free to break.
Client build
Communauté Coiffure
A full dashboard app, start to finish. If you want to see what a finished client project actually looks like, look here first.
Open sourceNodeGX
A visual app editor, rebuilt from the ground up with Claude Code. The clearest evidence I have that the teaching claim on screen three is real.
Proof of concept · free to tryNexus
An alternative to blockchain that signs with the phone already in your pocket, instead of a wallet you have to understand first. Go and play with it.
ConceptThe Data Spine
Bronze to silver to gold, borrowed from Databricks and pointed at the event industry: unusable data, to a single source of truth, to data that earns.
Things I've written
If you'd rather read than talk.
The guide
AI-Assisted Software Engineering
A methodology that actually works, written over many iterations. Brainstorm, document, execute, audit — how an inexperienced person ships production software without wrecking it.
ArticleIs your event data spineless?
Bronze-to-gold data consolidation explained for people who don't work in data and shouldn't have to.
Article · less seriousFor the woman and her tomatoes
How to actually start getting AI to fix your problems. Context is everything, and you are going to end up dependent. Welcome to the beautiful machine dystopia.
The long versionFrom zero to app development hero
GeoCities, spreadsheets nobody credited me for, a two-year project I abandoned, and what finally worked.
Playbook · not yet actionedThe GEO playbook
Generative Engine Optimisation — getting found by the machines people now ask instead of Google. A proof of concept, included because breadth is rather the point.
VideoSimple Rick tutorials
If you want to know whether I can explain things before you pay me to explain things, this is the cheapest way to find out.
The catch
Everything above is just me typing.
I could tell you I gave Steve Jobs the idea for the iPad. You can't check a word of it from here — which is why the only offer that matters is fifteen minutes of eye contact.
Book a call
Free, and shorter than you'd expect.