So how do you kind of self-regulate yourself on this?

I think what you need is a very good evaluation framework and so indeed we see different buckets in how we use AI.

There is the optimization of the work of every single employee that we work on and we build a lot of tools.

I think people underestimate how tools shape culture and how culture shape performance.

And so we invest a lot in building our own operating system and how to work at Alan.

We gained significant productivity.

And it's not about saving the software as a service cost.

What made you pick these three other countries? So what's the method behind the madness? And the second question is, how did the product travel? Were there some challenges? Have you sort of tweaked the product? Is it still the same?