Stage 01 · the 2000s
The bedroom
Your dad brings computers home, takes them apart, and puts them back together while you watch - hardware, though, always hardware. There's a family computer somewhere in the house, but software - websites, code, any of it - just isn't a thing that exists in your head yet; not because you considered it and passed, but because nobody has ever once put it in front of you. School never quite clicks. Sixth form doesn't either (a pattern forming, in hindsight). You make a proper hash of your exams, sit down with the university list to have an honest look, and find nothing on it - genuinely nothing - you can imagine caring about.
So you don't go.
The panels
You finish up the supermarket job and go to work for your dad - wiring electrical panels, on building sites at 6am, upgrading the office IT because you're the youngest person in the building and therefore, obviously, the computer one. You make a hell of a lot of teas and coffees. It isn't a tech career, or anything close to one. It's just where one starts, though nobody in the building would have guessed - you included.
The switch
One day you start poking at the company website - no plan, just poking - and some internal switch flicks on. This is the first time websites or code have ever been on your radar; nobody taught this at school, and nobody's teaching you now. The internet is. You seem to be - and it takes you a while to say this without wincing - quite good at it. Either way, you can't leave it alone.
Most decisions from here work the same way, honestly: it looks interesting, so you pour hours into it.
The first invoice
Somebody pays you for this. Actual money, for the thing you'd been doing instead of sleep - which changes what it is somehow, even though the work itself doesn't change at all. Nobody at school ever mentioned this as an option (a recurring theme, school not mentioning things). By January 2012 you're blogging; by summer you're telling other developers to say yes to everything, to hire yourself. Advice you'd been slowly taking for a while before you wrote any of it down.
A few hundred quid - not much, but enough for the hobby to quietly become a trade.
Stage 02 · 2012–2019
The workshop
You ship things, constantly, without ever really deciding to. KashFlow hires you in 2012 to rebuild their accounting UI; by 2014 you're out on your own for good - the freelance years, which is a tidy name for a decade of walking into other people's codebases (Kayako, Cisco, Plasso, Soho House, mytheresa, GoDaddy) and trying to leave each one a bit better than you found it, with mixed results you mostly never hear about. In between there are workshops, you teaching JavaScript to rooms full of strangers, and the side projects quietly piling up at the edges.
The shelf
Filtergram (Fast Company writes it up). Shades. Ekko (sold, 2020). Somna. Closeloop, ContentCaddy. Most of them die - of neglect, mostly, or of the next idea being shinier. Fine. Every one taught something the next one used, and anyway, you'd have built them regardless; that part was never really a choice.
The weekend
You build a fake API in a weekend, mostly so front-end developers have something to test against while they wait on the real one. You name it ReqRes, which is not a name anyone agonised over. You have no idea what you've just started - though to be fair, you could say that about everything on this desk.
Stage 03 · 2019–2024
The boardroom
You ask for a title nobody explains. Over email, in fact - and the reply is a mildly surprised “I didn't think you'd be up for that?”, which is fair, because neither did you. You ask because it'll force you into a growth you can't back out of; that's the theory, anyway, and it sounds better with hindsight than it felt at the time. The desk gets bigger and less yours. For the first time, the thing you're building is mostly made of people, and nobody hands you a manual for that either.
First seat
CTO of a bank that doesn't exist yet - engineer #1 of a platform that starts life as you making a directory on your MacBook called "bank" and thinking: I guess? Months later, everyone gathers on a call and the founder transfers $100 of his own money into the thing you've built. Six browser tabs, two terminal windows, a database GUI, and the longest thirty seconds of your career. Then he holds up the app: $100. No big climactic celebration - just relief, and quiet, solid proof that the system is real.
The platform swaps banking partners three times in eighteen months. You learn to build for the swap, there being no real alternative.
Second seat
The same job, and a completely different job. You join as a senior engineer; six months in, the founder reshuffles the team and you're in the seat again - quicker than you'd have chosen, if anyone had asked, which they didn't really. You know where the bodies are buried this time, mostly because you buried the first lot yourself. The pager on the table is face-up because of the weekend a DNS change took the whole API down while your alerts were switched off - a mistake you only make once, and one you got to make in front of everyone. The fix was better alerting, and it took an afternoon.
The notes
You wanted Kitchen Confidential for startups, went looking, and couldn't find it - apparently nobody writes the honest version while they still work in the kitchen. So you start writing down everything nobody told you: the 4am version, the version with your own name in the blame column. The whiteboard fills up faster than you'd like. None of it is a book yet - it isn't even trying to be one. That comes later, once you're out of the kitchen.
Stage 04 · 2026
The farm office
The desk faces a window now. It took a detour to get here - a year building billing rails at Clerk, back to being just an engineer on somebody else's roadmap, which turned out to be more of a relief than you'd like to admit. Now you consult from here - tech lead on a government data platform, which is not where anyone in this story thought the wired panels were heading - and you're building again too, because it turns out you never actually stop, you just change what you tell people you're doing. Seventeen years to get a desk with a view. The commute's better, anyway.
Still running
ReqRes, the fake API from that 2015 weekend, is still running. It outlived products you loved much more - which you try not to read too much into. It was never on a roadmap. It just keeps quietly going, mostly without you these days.
The weekend project now serves over 150,000,000 requests a month.
The book
The printed book standing on the desk is The First CTO: The Job Nobody Explains. It's the whiteboard from the boardroom, typed up and finished - everything you needed someone to tell you in week one, arriving several years too late to help you personally. Which is sort of the point.
now
The light stays on
Most things you build will die. You've made your peace with that - well, mostly. The switch that flicked on at your dad's company in 2009 never went off, and the livelihood it started is the one still paying for this desk. That's about it, honestly. It never really adds up to a moral, and you've stopped needing it to.