Tell me more about the operations for your AI.

Operations for AI applications is kind of fundamentally different than just regular applications.

Because with AI applications, tracing gets way more important to get...

And not only tracing, you just need a ton more of traces.

With classic applications, you might implement some tracing, like autotracing or whatnot, and you may sample your requests to get a good overview or a basic view of your deterministic operations in your application.

If you have an LLM doing the magic trick in your application, then you basically should trace 100% of your traces to get the full view because When using LLMs in applications, you open up your application to non-deterministic, to the entropy of the world, and it will do funny things.

I just had a talk and did a live demo, and it basically did funny things.

Yeah, I mean, you have to distinguish