What can $5,000 get you?
Real results in your repo by Friday.
Give me one week and your backlog. I’ll show you.
Most teams aren’t short of ideas. They’re short of someone who can say which ones matter, and then get them done.
I do both: I decide what’s worth building, and I run a development system that builds it while I keep deciding.
In one week you get
- 1
A verdict on your backlog
Every item in your backlog marked build, cut or defer, with a one-line reason. Most backlogs hold work that shouldn’t be built at all.
- 2
Real items, done
Three to five items from the top, built in your own code, each with tests that prove it works, and reviewed by me before it reaches you.
- 3
The numbers
What was built, how much passed its tests on the first try, how long it took, and what your team had estimated for the same work.
- 4
What a month would look like
If you like what you see, we keep going at a rate based on what the week showed. If not, you keep everything and we part friends.
No codebase yet? That’s fine.
Access to your code is the fastest start. If you’re building something new, user stories are enough: you get working software built from them, in a repository that’s yours.
Your code stays private
It is never sent to an AI company like OpenAI or Anthropic. The work runs on open models on my own machine. Nothing is used to train anyone’s AI.
Built this way
- WTF Tracker: from a written plan to a working app in 2 days. 73 tasks, 93% right on the first try.
- EstateWise: a rental-management app for our own use, not for sale. 421 tasks, built largely unattended, much of it overnight.
What happens step by step: how I build software →
One week. $5,000.
Half to start, half on delivery. The week starts once I have your code or your user stories.
Where is your backlog stuck?
Tell me about it in a few lines — an app, an internal tool, a codebase that needs work. I'll reply with how I'd approach it and a rough size.
Prefer to talk? Book a 30-minute call