Nvidia case study cookbook for x

I am also interested as well!

@JerePlum99 or @ChrisMG if you have an example of what you are trying to do with chainlit, that would help quite abit

(tbh, I've never used it LOL but looking at the docs, handling streaming and workflows should be straightforward)

I haven’t branched into workflows yet, if I come up with a usage internally here at my company I will share a notebook or something.

@Logan M you can see this example in the Chainlit cookbooks.

https://github.com/Chainlit/cookbook/pull/138

Right now, I’ve struggled with how to properly nest the parent / child steps for a single run into the Chainlit “steps” framework properly.

The above example differs from my core goal which is a custom function calling agent workflow that can stream back the async tool calls.

I agree it should be a straightforward use case and think a llamaindex/chainlit co-op would be super powerful for bootstrapping people.

Do you have the code for a specific worklow(s) that you are trying to use though? Having trouble picturing your setup

@Logan M yea I’ve been working on converting a legacy agent, let me clean stuff up this weekend and restructure as a reprex. I have proprietary data so need to make a different version