I was fifteen the first time I had cheesecake. Fell in love immediately — that dense, creamy texture, the graham cracker base, the way it somehow manages to be both rich and light at the same time. I ate three slices.
I was twenty the first time I wrote serious code. Fell in love with that too — the logic of it, the way a problem would resist you for hours and then suddenly give way. That dopamine hit when something finally works.
For most of my adult life I assumed baking was like coding: something that looked effortless from the outside and required years of practice to actually do. Not for people like me. I’m an engineer. I eat the cheesecake. I don’t bake it.
Then a few years ago, my wife found a recipe online — dead simple, fifteen ingredients, one video. She sent it to me on a Saturday morning with the message: try this.
I bought the ingredients, cleared the counter, and started.
Twenty minutes of prep took me closer to an hour. I was slow, careful, second-guessing every step. She helped me set the oven temperature. In went the batter. Thirty minutes to bake, eight hours to cool overnight — and what came out the next morning was one of the best cheesecakes I’ve ever eaten. Better than most I’ve bought. A fraction of the cost.
I’ve made it three times since. I’ve even attempted a Basque cheesecake. My son asks for it by name now.
I’m telling you this because you have an idea.
You’ve had it for a while, maybe. Something that should exist. A tool for your industry, a product for a problem you’ve seen up close, a service that could run itself if someone just built the right thing to manage it.
And you’ve been held back by one assumption: building it requires skills I don’t have.
That assumption was true for most of my career. Building software required developers. Developers were expensive, hard to hire, and spoke a different language. The gap between an idea and a working product was wide enough that most ideas stayed ideas.
Claude Code is the recipe.
You probably know enough
Here’s the bar you need to clear to use it effectively.
You understand that an Excel spreadsheet is just a form of database. You know “the cloud” means something runs on AWS or Azure rather than your laptop. You can look at a screen and identify what’s a textbox, what’s a dropdown, what’s a checkbox, what’s a button.
That’s it. That’s the prerequisite.
If you can describe what you want a screen to do — “when someone fills in this form and clicks Submit, I want it to send me an email with their details” — Claude Code can build it. You don’t write the code. You describe the outcome. It handles the implementation.
The skill you’re actually developing is specification: learning to say precisely what you want, and to recognise when the output doesn’t match what you meant. That develops fast. A few hours in, you’ll know what good output looks like and what bad output looks like. You’ll know how to push back.
What you can actually build
Start with the smallest version of your idea that would teach you something real.
Not the full product — the piece you could put in front of five people next week. A form that collects information. A dashboard that displays it. A simple tool that does one specific thing better than a spreadsheet does.
Claude Code takes that from description to working prototype faster than you would believe before you try it. From there, you iterate. From there, you show it to people. From there, the idea becomes a thing that exists — something you can point to, test, charge for, improve.
POC to MVP to first paying customer is a path that used to require a technical co-founder or six months of agency work. For a well-scoped idea with a founder who knows what they want, that path is now a few weeks.
My cheesecake didn’t need me to be a pastry chef. It needed me to follow a good recipe and pay attention to what came out of the oven.
If you’re the ideas person — the one who can see the problem clearly, who knows the industry, who knows exactly what the thing should do — you now have the recipe.
The only thing left is to decide to try.
I work with 0-to-1 founders who are building with AI tools and want a technical partner to validate the approach, stress-test the architecture, or take the product to the next level. Book a call — thirty minutes, no obligation.
