Oh, I harp on that too, but we'll get into what that means in a second.

Yeah, essentially, if you just break it down like a simple calculation, if you said, what's the max drawdown for Mag7 stocks? That would hit, we would translate that question.

We would pull that market data from our database.

Then that gets passed to a maximum drawdown calculation we've already defined and coded up.

And so we'll achieve that result.

That then gets passed up for the AI to summarize.

And so we really don't do any write the code on the fly approach or anything like that.

Give me a common use case you've seen that would not be viable or would require a lot of swivel chair action or hours of work that you can now enable within minutes that you've seen people commonly use.