You Can Build Your App With Codex — Even If You're Not a Programmer

What if you already know enough to start building something useful?

Not because you know how to code.
Because you know your work, your business, your problems, or your process.

That is where my experiment started.

I had a real business problem. Creating waterproofing estimates involved calculations, product requirements, labor, project conditions and rules I already understood from experience.

I didn't start by saying:

“I want to become a programmer.”

I wanted to solve a problem.
And that eventually became an app.

I Had the Knowledge. I Didn't Have the Programming Skills.

That difference became one of the biggest lessons from this project.

I knew how my business worked.
I knew what information I needed from a customer.
I knew many of the rules behind creating an estimate.

What I didn't know was how to turn all of that into software.

That's where ChatGPT and Codex entered the story.

ChatGPT helped me think through problems, organize requirements, understand errors, review results and decide what needed to happen next.

Codex helped work with the actual project—implementing changes, working with code and files, testing and troubleshooting.

And I stayed in the middle.

My knowledge + ChatGPT + Codex + testing + corrections.

It wasn't one magic prompt.

And Yes, Things Went Wrong

This is an important part of the book.

AI didn't magically build everything correctly the first time.

There were mistakes.
There were things I explained incorrectly.
There were things AI got wrong.

We discovered security concerns that changed how parts of the application needed to work.

We restored from backups.
We tested.
We corrected.

At one point, Codex itself stopped working correctly.

Instead of pretending that didn't happen, it became another lesson.

I used ChatGPT to help understand what Codex was telling me and troubleshoot the problem step by step.

That experience produced one of my favorite principles from the book:

You don't have to understand every technical word in an error message to begin investigating the problem.

This Book Isn't Really About Waterproofing

The Waterproofing Estimator is the case study.

But the bigger question is:

What do YOU already know?

Maybe you understand:

• Construction
• Accounting
• Real estate
• Cleaning
• Landscaping
• Automotive work
• Sales
• Scheduling
• Fitness
• Food
• Property management
• A hobby
• A repetitive process at your job

You may know a process that somebody else doesn't.

That knowledge could potentially become a calculator, checklist, estimator, client-intake tool, reporting system, simple app, automation, digital product or service.

That's what I mean by transferable skills.

Don't just copy what I built.
Learn how I approached building it and apply those lessons somewhere else.

What You'll Learn in the Ebook

You Can Build Your App With Codex follows the actual journey from business problem to working application.

It covers identifying a useful problem, explaining your knowledge to AI, working with ChatGPT and Codex, improving instructions, testing what AI produces, correcting your own mistakes, protecting sensitive information, keeping backups, troubleshooting problems, and gradually turning an experiment into something useful.

And it does this from the perspective of someone who didn't start as a programmer.

There are no promises that AI makes everything easy.
It doesn't.

But AI can dramatically change what you're capable of attempting.

Start With What You Know

You don't need to begin with a revolutionary startup idea.

Look at your own work.

What do you calculate repeatedly?
What information do you collect over and over?
What takes too long?
What do customers constantly ask you?
What process exists mostly in your head?
What do you know how to do that took years to learn?

There might be something there worth building.

Start with what you have. Learn what you need. Build something useful.