vibecoding, app development, ai-tools,

Introduction

Why am I telling this story?

Because I'm often asked how it's possible today to build a functioning app in just two weeks , even without a developer background - with the right idea, a lot of focus and the support of generative AI .

What should not be underestimated:

• Over the past 20 years, I've tried every new programming language - but never had the focus or skills to write more than a Hello World app.

• I am a business information systems specialist and, together with Oli De Bastiani, I developed an app that generated over 500,000 downloads .

• I invented the first wooden payment card without a plastic core and scaled it globally with the industry.

In short, I understand a lot about product development and what it takes to turn an idea into a product.

But what I've learned in the last year opens up unimagined possibilities for developing and marketing digital products. I'd like to share this story in the coming weeks.

The story is titled:

“Becoming Mr. Smith – My journey as a vibe coder” .

Mr. Smith—aka Bionic Bedo —is the name of my AI agent, which is primarily based on Claude AI. However, I use a toolset to develop my apps like DinBrudi .

And in the future? There won't be just one Mr. Smith, but entire teams of agents (my Team48)—each specializing in a specific area.

And like every story, this one begins with a decent breakfast or a beer.

From conversation to vision

Prototyp von DinBrudi

In September 2024, I sat with Christof Zogg for a long breakfast. As always at

Our conversations were brimming with ideas:

👉 Who else should we talk to?

👉 Which new technologies should we definitely look at?

Naturally, the topic of artificial intelligence was also very present. But at the end, something typical happened: Neither of us knew exactly what we had agreed to or which topics we specifically wanted to explore.

I noticed that this problem occurs not only in such conversations, but also in workshops and meetings within the company. There is often a lack of clear conclusions, responsibilities, and next steps.

At that time, there had already been initial attempts at conversation transcriptions in Teams , but I wanted a solution for physical meetings . This is how the idea for DinBrudi was born:

💡 An app that listens, summarizes, and creates a to-do list.

A month later, I built a first, very small prototype, which was still called Bedo's Alfred, because who doesn't want to be Batman ;-)

As I said, I'm not a coder , but I can use development environments like Xcode or Android Studio . The results were surprisingly promising—there was only one problem: Swiss German wasn't understood.

In the next part, you'll learn how I built a product for the App Store in two weeks of vacation.