I’ve just finished another small project.

This time, it’s a website where you can search for your favorite band and see when they last played in your area. The idea is simple, but useful—especially if you want to nudge your favorite artists (particularly those from the US) to include Northern Europe in their future tour plans.

The Idea

The app lets you:

• Search for an artist or band
• See where and when they last performed
• Use that information as a gentle argument: “Hey, it’s been a while since you were up here.”

The concept itself isn’t new. It’s actually a clone of a project a friend of mine coded last year. This time, the experiment wasn’t about originality—it was about speed, tooling, and execution with AI.

Built with Claude Code (Very Fast)

Claude Code built the initial version of the app in
3 minutes and 34 seconds.

That first version:

• Mostly worked
• Contained one bug
• Was missing one feature I wanted

I fixed the bug and added the missing feature myself. That part still mattered. Even with AI moving fast, human judgment and small manual tweaks made the difference between “it runs” and “it’s usable.”

The App Is Live

The site is now live here:

👉 https://bltg85.github.io/last-gig-finder/

To actually see results, you’ll need to provide your own API key. Without it, the app won’t be able to fetch live data, and you’ll end up staring at an empty result state instead.

A Small Reflection

What stood out to me in this project wasn’t the complexity—it was the contrast.

• An idea that took a human weeks last year
• Rebuilt in minutes with AI
• Still improved by human context and intent

This is quickly becoming the new normal:
AI gets you to “something” incredibly fast.
You still decide whether that something is worth shipping.

And sometimes, that’s all a project needs to be.