HOW EVERY BUILD IS PROTECTED
Anyone can build it now. Almost nobody builds it right.
AI changed what “working software” looks like. Anyone can ship something that demos well — you can get a working demo in an afternoon, and so can every developer you’re evaluating. A demo is no longer proof of competence.
The difference between a demo and a business system is everything that happens when nobody’s watching: the error at 2am, the API that silently changed, the database that needs restoring, the invoice email that stopped arriving.
I build the demo fast too. Then I build everything on this page — because your business runs on it. It’s the same reason the process starts with a working prototype before you pay a dollar: the prototype is proof I understood your business. This page is proof the build survives contact with the real world.
The rule for this page: a claim that can’t point to its enforcing mechanism doesn’t belong on it. Everything below says what it is and how it’s enforced.
THE CHECKLIST
The questions nobody asked you
If you’ve had a system built before — or built one yourself — read these as a checklist. Your last developer should have had answers. Most never heard the questions.
What happens when something errors at 2am?
Every error is captured and reported the moment it happens — nothing fails silently. On the care plan, those alerts reach me directly, by phone — I usually know before you do.
How do you know it's even up?
Automated checks test the system from the outside, around the clock. "We assume it's fine" is not monitoring.
When did you last test restoring a backup?
Continuous backups plus separate offsite copies — and restores get tested. A backup you've never restored is a hope, not a backup.
Who's watching your domain and email?
Automated checks watch your domain records and your email deliverability around the clock — because nobody thinks about the domain until the site vanishes, or deliverability until the invoices stop arriving.
Where does customer data live, and who can see it?
Encrypted, access-controlled at the database level, and private information is filtered out of logs and AI calls by design.
Does it fix itself?
Where it's safe, yes — automatic retries and recovery for transient failures. Where it's not safe, it alerts a human (me) instead of guessing. Software that edits itself unsupervised is how the same bug gets "fixed" five times.
What do you actually own?
Everything. Your accounts, your domain, your data, your code — from day one. If we parted ways tomorrow, you'd lose nothing but me.
If your developer disappeared tomorrow, could anyone take over?
Yes — because every build ships with the Owner's Manual. Keep reading.
THE PATTERNS
The three ways tech guys go bad
Business owners have been telling me the same three stories for as long as I’ve done this work. If one of them is yours, you’ll recognize it by the second sentence. And notice what’s missing below: the word “ethical.” Claiming it is worthless — the last guy’s website claimed it too. What matters is the mechanism that makes each pattern impossible here.
PATTERN_01 — THE PROMPT JOCKEY
“Done” without testing.
He takes your bug report, pastes it into an AI coding tool, the tool says "done," and he tells you it's fixed — without ever running it. Next week: same bug. Your best person burns billable hours re-reporting problems that were never actually fixed. AI made this guy ten times more common, because the tools are now good enough to look done.
Every fix ships with verification — what changed, how it was tested, what confirmed it. Automated tests and monitoring catch regressions before you do. You hear "fixed, and here's how I know." Never just "fixed."
PATTERN_02 — THE HOSTAGE-TAKER
Holds your business's keys.
He registered your domain in his account. He hosts everything under his logins. He charges monthly for "maintenance" that's mostly nothing — and you can't leave, because leaving means losing your website, your email, sometimes your whole online presence. Only he knows where anything is. Of the three, this is the story I've heard most.
Everything is registered in your name from day one — domain, hosting, data, payments, code. Hard rule, no exceptions. Built on your infrastructure, not mine with a login you borrow. And the Owner's Manual means any competent developer could take over tomorrow.
I make it easy to leave me. That’s why nobody does.
PATTERN_03 — THE DOUBLE-DEALER
Sells your tool out from under you.
The extreme case — and it really happens: he builds your tool with your money, then markets it to your competitors. Sometimes he rewrites the deal mid-project and starts charging you monthly for access to the thing you already paid to build.
The IP terms are in the proposal before you ask. What I build for you is yours. I reuse patterns and experience across clients — never your tool, never your data, never with your competitors. In writing, unprompted.
The honest version of “clients come first”: my clients stay for years, and not one of them is locked in. Retention by results, not by ransom.
THE DELIVERABLE
Every build ships with the Owner’s Manual
The documentation your last developer never gave you. It’s a real packet, handed over at launch — written for you, the owner, not for a developer.
SYSTEM MAP
What the pieces are and how they fit together. One diagram, plain language.
OWNERSHIP LEDGER
Every account, service, domain, and repo: what it is, who owns it (you), where to log in, what it costs.
THE RUNBOOK
What happens when something goes down: what's automated, who gets alerted, and what you should do — usually nothing, because here's who's already on it.
BACKUP & RECOVERY
What's backed up, where, and when a restore was last tested.
SECURITY OVERVIEW
How your data is protected and who can access what.
INTEGRATION MAP
Every external connection, what flows where, and what breaks if a vendor changes something.
OPERATING GUIDE
How your team actually uses it, per role, with the recorded training from launch.
CONTINUITY PLAN
Exactly what happens — and who can take over — if I'm unavailable. The real "hit by a bus" answer, in writing.
It also says plainly what you don’t need at your size — and the trigger that means you’ve grown into needing it.
If we parted ways tomorrow, any competent developer could pick this up before lunch. And if you’ve been burned before, here’s a test that costs you one email: ask your last developer for any one of these documents.
NO HANDOFFS
Who you’re actually working with
Me. Jon Barrett — the person who reads your application, builds your prototype, writes the code, and answers when something breaks. No account managers relaying messages, no dev team you’ve never met, no offshore mystery behind the curtain.
Small is the point. Nothing gets lost in a handoff because there isn’t one — the person who built your system is the person watching it run.
AFTER LAUNCH
The opposite of rent
$0
PER MONTH · FIRST 90 DAYS INCLUDED · YOU OWN EVERYTHING EITHER WAY
Some vendors charge you monthly to access your own software — rent, on a thing you already paid to build. This is the opposite, twice over. First: the monitoring on this page is built into every system I ship — instrumented, alerting, and backed up from day one, whether or not you ever pay me another dollar. Second: what the care plan buys is me on the other end of those alerts — hosting, fixes, small improvements, and a direct line, with every change logged in your change log so you can see exactly what the monthly fee bought.
The first 90 days are included with every build. After that it’s optional, because the Owner’s Manual means you could run it yourself or hand it to anyone — the alerts and the runbook work the same for whoever answers them. Most clients keep it. Not because they’re stuck — because it works.
If you just realized your system has none of this…
…that’s fixable, and you’re in good company. You built something that works — nobody told you about the part after “works.” Apply, tell me what’s running unprotected, and I’ll tell you what it needs. If it’s small, I’ll tell you that too.
Apply to work with me →QUESTIONS