In my workflow, I set the conditional type to be 'AI', and the description is to go down the path if the prospect agrees, and another response for when the customer disagrees. During the conversation, the voice agent just continues down a set path without waiting for response. Would someone be able to provide me some insights? I remember that I don't need to use the 'gather' function because I'm not looking for a specific variable to store. Call ID: 070e4655-de5b-4081-b8a7-4a2bc0ef8891 742ce08b-5a8f-445c-a6e7-f82b5dc1bdb8 b0ab0818-9e20-47c2-a4ad-243eea6dcfee a25fa946-9355-4a69-a780-4a165096e72d

Thanks for sharing the details of your workflow issue. Based on what you've described, it sounds like the voice agent isn't branching as expected when using the 'AI' conditional type, and it's not waiting for a response before proceeding. Unfortunately, I don't have specific information about how the 'AI' conditional type should be configured or whether the 'gather' function is necessary in this scenario. If you have access to more details or documentation about the workflow setup, that might help clarify the expected behavior.

Add a Gather node after your prompt to ensure the assistant waits for user input before branching based on AI conditionals, even if you’re not storing a specific variable.

I've used the gather function, but it specifically asks for inputs from the user, e.g. "Can you confirm that _______" Is there a way I can remove this? It breaks immersion for the user. Also, after using the gather function, the conversation does not flow when it is linked to an assistant. Could you look into this issue? Call Log: 5041167f-b5c5-440a-8d62-b5675f4bb184 4a92f28d-1597-4e82-94a8-8613799d80b4 904d7407-273b-4836-aefc-faa55843c58c

Looking into it

Hi, has there been a resolution?

You can also use the conversation node to gather info where you can write a prompt to get the information required in a way that suits your needs. Let us know if you have any questions regarding that

I do, but as I mentioned the agent will ask for the same input from the user, breaking immersion. There also seems to be an issue with assistants in workflow, please refer to the call logs below: fd128226-529e-4795-9ceb-5ce9223b93f2 c737992d-1715-4105-9901-2f315e03ef4f 73381963-2b8b-4d43-939b-465ade5e927c

We are currently in the process of revamping workflows and we recommend temporarily moving to single-prompt assistants with tools for the time being.

Sure thing. Is there a timeline I can look towards?

Hey, we don't have exact timeline to share but you will see the updates soon.