I’ll be honest from the start: MadPlanner is built by AI. But not in the way the phrase is often used.
There’s an idea that “built by AI” means something thrown together over a weekend — a quick prototype that looks good but falls apart the moment you touch it. This is the opposite.
A year, not a weekend
MadPlanner has been developed step by step over more than a year, with AI as the throughline engine — from the code to the setup, from deployment to the daily operation. I’ve used nearly every tool out there, tested them against each other, and built something that’s engineered, not just generated.
It’s my personal project — one of several — and the first time I’ve tried something at this scale. It’s been a long learning curve, and it’s far from over.
Why I think it’s the future
I believe we’re heading towards a world where AI doesn’t just write snippets of code, but builds and maintains the whole product: sets things up, deploys, fixes bugs and runs a small business, step by step. MadPlanner is my attempt to build exactly that kind of product — in the open, while I learn it myself.
That also means I’m not promising perfection. We’re still in beta. We’re testing, improving and developing all the time, and the roadmap is still long. Some things don’t work perfectly yet — and that’s part of building something real instead of pretending it’s all finished.
What I’m hoping for
The goal is simple: for MadPlanner to bring real value to families in a busy everyday life. Not an impressive AI demo, but a tool that actually makes dinner easier, week after week.
If you try it and something grates, that’s exactly the kind of thing I’d love to hear. We’re building this together with the first families — that’s how it gets better.
Try a free meal plan and see where we’ve got to.