So I need to use 4 tools within a call, and report the results of all 4 in the end-of-call-report. Then I will map these 4 fields into a googlesheet row along with the transcript, summary and other standard end-of-call report stuff. I see some previous bot answers about using a multi_tool_use.parallel function, but have a few questions on setting this up. 1. Is this pre-existing in the functions tab or the advanced tab (searched and didn't see it in either), somewhere else, or do we have to create it as a new tool? 2. Please tell me step by step by step how to create this, I've been trying way to long to use multiple tools in one call, please break down every single step thank you

@Pennystarfire Absolutely, I'd be happy to guide you through the process of using multiple tools within a call using the

multi_tool_use.parallel

function. Here's how you can set it up: 1. **Tool Availability**: The

multi_tool_use.parallel

function is a part of the

multi_tool_use

namespace, which is designed to run multiple tools simultaneously. This function should already exist within the VAPI system, and you don't need to create it as a new tool. It's not located under the functions or advanced tabs but is part of the API's capabilities. 2. **Step-by-Step Guide**: - **Step 1**: Identify the tools you want to use. Make sure each tool is compatible with parallel execution and is listed under the

functions

namespace. - **Step 2**: Gather the necessary parameters for each tool. Each tool might require different parameters, so ensure you have the correct information for each one. - **Step 3**: Use the

multi_tool_use.parallel

function to execute the tools. You will need to pass an array of objects to the

tool_uses

parameter, with each object containing the

recipient_name

(the name of the tool) and the

parameters

(the parameters required for that tool). - **Step 4**: Call the

multi_tool_use.parallel

function with the array you've prepared. The function will run all the specified tools in parallel. - **Step 5**: Once the tools have been executed, the results will be returned to you. You can then compile these results into your end-of-call report. - **Step 6**: Map the fields from the results, along with the transcript, summary, and other standard end-of-call report information, into a Google Sheets row. If you provide me with the names of the tools you wish to use and their respective parameters, I can help you construct the call to the

multi_tool_use.parallel

function.

sure, here are the 4 tools and their parameters. 1. confirmReceivedEmail: If the customer confirmed they received the appointment booking email, result is 'Yes'. If they deny, result is 'No'. 2. confirmAttendAppointment: If the customer confirmed they will be attending the appointment, result is 'Yes'. If they deny, result is 'No'. 3. getInterestLevel: If the customer is interested in our done for you offer, result is 'Yes'. If they are not, result is 'No'. 4. callAnswered: If the customer speaks, result is 'Yes'. If they do not say anything, result is 'No'. Thank you @User

You just need to mention it in prompt and model is smart enough to take care of it

Through the dashboard prompt? Or do you recommend everyone just use the code based, cause I see a lot of the help is to alter the code-based create assistant method, update using the code, etc. I wasn't sure if the support bots step by step guide was dashboard based or code setup based either, or referring to automations modules using arrays etc. But yeah I tried putting it on the dashboard prompt along with what tools and parameters, and it still only pulled through one of the message, type, toolcalls, functions, arguments strings with a value of 'yes', as seen on the make webhook output. The rest of tools were only ending in a description of the tool rather than argument value 'yes' or 'no' which I need to map into the row. Row values include phone number, name, etc, etc, answered 'yes' or 'no', received email y/n, will attend appt y/n, interested dfy offer y/n, summary, transcript, etc. as headers and info being pulled in so just you have an idea what I mean. Thank you

It is simple prompt change you can do it via dashboard or code