Sup everyone! Im using the Vapi web UI and need some help figuring out the best way to do one of two things: a.) Collect multiple response options from the person being called, or b.) Let the AI determine whether certain phrases I’ve defined match what the person is saying. I'm building a feedback assistant that follows a dialogue tree based on logic, but I’m running into serious trouble setting it up. The goal for my assistant would be to call customers, ask for a specific person, and then follow different branches of the conversation depending on their response.

To collect multiple response options use structured prompts and decision trees, by setting trigger phrases and using entity recognition, enabling your assistant to branch conversations based on customer replies..

Thanks for your reply! I’m still quite new to VAPI and just getting familiar with it, so I don’t fully understand this yet. Would I need to code my assistant manually instead of configuring it through the Web UI? The screenshot shows how I’ve set it up so far, but I’m assuming it’s not quite right and far more complex? https://cdn.discordapp.com/attachments/1362042216188739727/1362049967564259579/swlIqC3.png?ex=6800fb61&is=67ffa9e1&hm=bf2218156f2d8224f721ebf42d577a9f6c6597f51f3baccd29909c387dd08763&

You can set up your Vapi assistant entirely through the Web UI by defining system prompts, initial messages, and conversation branches using keywords or phrase matching, no coding needed unless you want advanced custom logic.

Thanks for your answer, what would be the best way to do that?

You can build an assistant via the Web UI by setting a clear system prompt, choosing who speaks first, and using workflows with keyword-based branches