I have some AI transitions and have changed the wording in the categories a few times to try and make it transition more accurately. I currently have it at the best it has been, but there are still some inputs that it transitions incorrectly when I test it. Is there a way to give these as examples and tell the AI what the correct transition was? Or is there any advice on what is best practice for improving AI transition accuracy?

I think the only way to train them would be to use Intents instead. It would be really nice though, if we could add examples like in the AI Task card!

do post this in #1111026806254993459

Done 🫱🏼‍🫲🏽

This would be great. I'm having the same kind of issues with the AI Transition. having examples like in the AI Task would be really cool!

Lets hope its not too far away! Not sure how it works on the backend but I'm thinking that because they have examples in the AI Task it might be relatively quick to implement, so fingers crossed 🤞 The key will be that they make sure the examples do not over ride the main information in the categories, it should be a fall back method when the AI can't confident classify the input