I stopped using travel apps — and built my own instead.

During my recent trip to Thailand, I didn’t use any travel apps.

Instead, I used Claude to build a fully bespoke tool for myself — covering everything from itinerary, mapping, checklists, journaling, photo feeds, all the way to a final trip summary.

This wasn’t a “one prompt and done” situation. AI is not magic. It’s a co-pilot that helps remove the biggest constraints: technical skills, coding, and complex logic — things that would normally take a lot of time to learn or require hiring someone.

A concrete example:

A friend wanted to visualize over 600 photos, mapped to the locations we visited — all at once on a map.

If you do it naively, it becomes a cluttered mess.

So instead of jumping straight into code, I started brainstorming with AI — exploring different representations: stack, network, spider web…

At some point, it became clear that the problem wasn’t about “showing images”. It was about how to organize space and information.

The final approach:

• Connect visited locations into a primary network layer

• Distribute images into the empty spaces not occupied by the network

• Use “lead lines” to link images back to their corresponding locations

The result:

• Images still anchor to their locations

• The density of the dataset is preserved

• But the overall visualization remains clean and readable

(Screenshots attached below.)

What’s interesting is:

AI didn’t come up with this solution on its own. It only became powerful once I had an idea that resembled an algorithm.

At that point, AI could translate the idea into implementation, refine the logic, and generate code that actually matches the intent.

AI only gives you real leverage when you’re clear about what you want.

• What are you trying to build?

• What problem are you solving?

• What value are you aiming for?

With today’s AI, we now have the freedom to build our own tools — personal, flexible, and at a fraction of the cost and effort before.