ship it · one to one · three places at a time

Your app, live
in four weeks.

Q: "Do I need to be able to code?"

"No. You need an app you want and about two hours a week. Claude Code writes most of it; I show you what each part does as we go, so by week four you can change it."

Q: "What fits in four weeks?"

"A first version of the app you want to build, with its main job working, live and safe for real users. If your idea is bigger than four weeks, we agree in week one which part ships first and the rest is yours to build with the same tools."

Q: "Do I get the skill library?"

"Yes. It is how we build in week one, and it is yours to keep using afterwards."

Q: "Why would you teach me not to need you?"

"Because that is the honest version of the help. The alternative is selling a dependency and calling it a service. It does mean this business runs on new people and referrals rather than on keeping anyone stuck, which is a real trade-off and I would rather name it than hide it."

Q: "Is this a course?"

"No. There are no videos. It is your app, built with you, with written feedback at each milestone."

Q: "What if I would rather you just built it?"

"Say so on your application. I will tell you straight whether that is something I can help with."

Four weeks, one to one, mostly in writing. You bring the app you have started with AI, or the idea you have not. We take it to live together: logins, backups, hosting, the store. At the end it is in real people's hands, it is yours, and you know how to change it.

A$1,500 · the first three places are A$750 in exchange for a written case study

Apply for a place

who this is for

It works on your laptop.

You have an app you want to build, and you have started it with AI or are about to. Getting it working was the fun part. Getting it onto other people's phones, safely, is where it stalled.

If you want something built and looked after for you indefinitely, without learning how it works, this is the wrong offer, and I would rather say so now. Say that on your application and I will tell you straight whether I can help another way.

It only runs while your laptop is open.
You are not sure whose data it holds, or who could reach it.
You know "put a login on it" is a sentence, not a plan.
You want to be able to change it yourself next year, not pay someone every time.

the four weeks

One milestone
each week.

Each week ends with something you can check with a yes or a no. If it is a no, we do not move on.

Planweek one

What the app does, who uses it, what it stores and who is allowed to see it, set up with the skill library. Milestone: a one-page plan and a task list you can read.

Buildweek two

The screens that matter, working with real data on your phone. Milestone: you use it for the job it is for.

Make it safeweek three

Logins, who can see what, backups, and error reporting. Milestone: a friend signs in on their own phone and uses it.

Shipweek four

Live somewhere it stays up, after a production check of security, backups and running cost. Milestone: it is live, and you have written notes to change it yourself.

Between milestones I review your progress in writing within one working day. No standing calls.

what it costs

One price, stated.

The whole four weeks, one number. The tool's running cost is separate and small, and you see it before anything goes live.

price

A$1,500

first three places

A$750, for a written case study you approve

places

three at a time

running cost

yours, stated before launch

If the week-four milestone is not met because of something on my side, I keep working with you until it is, at no extra cost. That does not cover things added after week one.
It is deliberately finite. A model that leaves you independent cannot also be a subscription you never escape.

what happens next

If you apply.

Four steps from an email to week one.

01
You email me today
What you are building, who it is for, how far you have got, and whether you want to learn to run it yourself or would rather it was done for you. A few sentences is plenty.
02
One conversation within a week
One call to check the app fits in four weeks and that I am the right person for it. If it does not fit, I will say so on the call.
03
Invoice, then week one the following Monday
Week one starts the Monday after the invoice is paid.
04
The handoff end of week four
One more call, and written notes, so the next change is yours to make.

why me

I build tools for
a market of one.

I am a developer by trade, and each of these runs today. None of them needed a team or a vendor, which is the size of app this is about.

productive-mcp, published on npm since November 2025 and installed about 100 times a month (101 in the month to 4 Oct 2026) by people I will never meet.
Gloaty, real infrastructure for a real group of friends.

the awkward questions

Asked and answered.

What I would want to know before applying.

No. You need an app you want and about two hours a week. Claude Code writes most of it; I show you what each part does as we go, so by week four you can change it.

A first version of the app you want to build, with its main job working, live and safe for real users. If your idea is bigger than four weeks, we agree in week one which part ships first and the rest is yours to build with the same tools.

Yes. It is how we build in week one, and it is yours to keep using afterwards.

Because that is the honest version of the help. The alternative is selling a dependency and calling it a service. It does mean this business runs on new people and referrals rather than on keeping anyone stuck, which is a real trade-off and I would rather name it than hide it.

No. There are no videos. It is your app, built with you, with written feedback at each milestone.

Say so on your application. I will tell you straight whether that is something I can help with.

apply

Three places at a time.

Email me what you are building, who it is for, how far you have got, and whether you want to learn to run it or have it done for you. I reply to every application.