Karl Yeh

for some weird reason.

So yeah, the, the data problem is there.

There is ways to do it, but that's why it's like, and then the extreme cases, Oh, we'll just use a proprietary system or like, um, an on-premise system.

Like, yeah.

Okay.

Do you have the resources, the capabilities and, um,

the ability to maintain all that and the GPUs because it's like, well, it isn't just five people using it.

There's going to be lots of people that you want to use it.

So there's different problems with different situations.

So it's like,

And that's why it's so unique to each company that you can't just be like, well, this is the data problem and we're going to solve it.

No, you've got to actually walk it through how they...

use their data.

And I think like the message that we always come out with now is there is no AI.

It's not what is the best AI it's what's the best AI for this specific project use case in this business context with your specific data, that could be any different things, which is, it just, it is just what it is today.

Yeah.

Yeah, it's... And it's... I think, like, it's tough because, like, majority of our clients are Microsoft clients.

So it's like, hey, what are you currently... Like, you're in the Microsoft environment.

So...

That and then you start there, but then you kind of start seeing because it's each each software they use is so the either.