How AI is changing the way we prototype software

So maybe...

You said, okay, it was pretty fast.

That is maybe also what we need to zoom in on, right? Because a prototype doesn't mean a picture perfect, pixel perfect, fully developed application.

It is very much that, right? Prototype.

What makes a good prototype for you? What makes it effective?

First of all, I think the prototype needs to deliver exactly what it was requested.

So, I mean, for example, you have, I don't know, a customer management software.

You can tell AI, hey, build me a customer management grid.

And if you just tell it that, it will build you a grid with all the Pampos functionality, search, inline search, I don't know, drag and drop management, stuff like this, which you don't need.

Yeah,

it starts

to

act as a product manager with a blackjack, right?

Exactly, exactly.

And then, I mean, you don't want that.

You just want to prototype it.

You just need to emphasize the fundamental business flows.

And then after the client agrees, you can then decorate it with other functionality.

I mean, any functionality that you add on top is converted in additional hours, development hours.

So, I mean, as a business, you need to decide, hey, do I want to invest this amount of money in the live search? Will it have any value for the customer? I mean, this is why you need to follow the KISS rule.

Keep it

simple.

And then, yeah, just stick to what was requested.

And I think that's the most important thing to think about

when you're prototyping.