START CALL – How to start a conversational flow scenario?
Last updated
Last updated
Every voicebot scenario starts with the Start Call block. And this is the only part of the scenario when this block can be used. That is why you will not find it in our Blocks Menu.By default, it will be placed in the Conversation Flow Scenario.
Start Call is the only block without the input port. It has only one output port.
Welcome text – here you may provide your greeting text which your voicebot will speak when connected to your client. This functionality works exactly as the Text to say option in the Say block.
Voicebot language - You can configure your voicebot in English, Polish, Spanish, and Hungarian. However, if you set it up in the language other than the language in which your texts are written, the voicebot speech will be incomprehensible.
REMEMBER! Before you start testing, check what language the voicebot is set to!
First API call URL - There you provide a Uniform Resource Locator (URL) to your Application Programming Interface (API) from which the voicebot can retrieve the data needed for a call. This allows for personalization. For example, to start an Outbound Campaign you need to retrieve the customer’s name and other necessary data from the system. Depending on the functionality of your voicebot, this may be, among other things, the amount of money the customer owes the bank, the date of the appointment, etc.
TIP: We require your API to return data in a specific JSON structure to set Global Variables.
Voicebot initial timeout – the time that the voicebot waits for the customer response.
Voicebot final timeout – the amount of time that the voicebot waits after the customer has finished speaking.
CSV file name - in case your voicebot uses a CSV file to exchange the data, specify the name of this file.
CLUSTER - Each InteliWISE customer owns one or more databases with the possible answers to bot questions. These databases can be divided in CLUSTERS that store a specific subset of answers. For example, you may want, different answers for a voicebot and different answers for a chatbot. You can also have several subsets for the same type of bot. For example, you may prefer different answers from a chatbot on Facebook Messenger and different on your company website.
Force ASR tolerance – while the ASR system recognizes speech to text – you may increase its tolerance. This means, that its ability to recognize words spoken by a customer will be higher.
Techmo synthesis – in case you choose the Techmo speech synthesis instead of Google synthesis.
Custom JS – here you can provide JavaScript code to do more advanced things. Usually, in this field you may set up global variables, choose the google voice, and configure some other basic settings needed for the whole conversation.
DNN confirmation detection - helps to force understanding of the customer’s speech in the case when a given block only expects a “Yes” or “No” answer. Probably in other cases, the voicebot might understand another word, but here the assumption is clear. This option is helpful in cases where the customer diction is not perfect and then we avoid bot failure. If this option is set in the Start Call block, it is enabled in all other blocks.
