Hi everyone.

We have some doubts related to Architect and how to play promts that depend on variables values without using “Text To Speech”.

On our IVR we need to announce dates or prices which are variables, and for this we have a prompt database which we use to concatenate these depending of variables values.

For instance, to announce a date like “November 13, 2017” (variable like “2017-11-13”) we need to concatenate five promts “November” + “13” + “2” + “thousand” + “17” and play it.

Our question:

How can we configure this on architect? Using a “Switch action” for each possible variable number? Could it be configured on “Play Audio action”?

Or is there a better way to configure it?

Thank you in advance.

Best regards,

Ana Laia

Hi Ana,

Take a look at the ToAudioDate function in Architect. There are multiple overloads to allow to specify how you want a date to be played. I think you'll want this one: ToAudioDate(MakeDateTime("2017", "11", "13"), Format.Date.longYear).

I'm not sure where your dates will be coming from, but you'll probably need to convert a string "2017-11-13" into a datetime variable; I'm attaching a screenshot showing how.

First I split the string up into a string collection, then I pull each piece out and convert it an integer, lastly I constructed a datetime variable from the individual parts. The expression for Task.date is cut off, so here's the complete value: MakeDateTime(Task.year, Task.month, Task.day)

For the prices, use the ToAudioCurrency function.

It sounds like you have your own recordings; for your customers to hear them instead of the system defaults, you'll need to overload the system prompts with your recordings.

https://help.mypurecloud.com/articles/access-functions-help/

https://help.mypurecloud.com/articles/system-prompts/

Hi Melissa,

Thank you for your answer.

Yes, you are right. On our contact center we will use our own recording database (in European Portuguese). Thus, we will import our prompts to Purecloud.

From what I understood, the functions “ToAudioDate” or “ToAudioCurrency” use some kind of Text to Speech. For instance, on their documentation it’s stated “Note: This assumes the flow's current language is English (United States) [en-US]”.

Therefore I think they won’t accomplish what I was searching…

On our recording database we have a prompt for each digit, for the currency (“Euros”), for countries (“Portugal”, “Spain”…..) etc.

In addition, we are receiving these values from an external system through data dip action.

Thus, the idea is to select the correct prompt to the value. If It corresponds to “2” I need to select to play the prompt “Dois” or if the country is “Belgium” I need to select to play the prompt “Bélgica”.

Can I add some logic on “Play Audio Action”? A “Switch”? Is there other way for doing this?

Sorry if this sounds a bit confusing...

Best regards,

Ana Laia

Hi Ana,

The functions don't use text to speech, they play the appropriate system prompts for the currently executing language of the flow. The note in the the expression help was explaining that the example audio is what is heard when the current language of the flow is en-US.

So when the flow is using en-US, the caller would hear the following system prompts:

daysofweek_Monday ("Monday")

month_november ("November")

num_13_ordinal ("thirteenth")

num_2 ("two")

num_thousand ("thousand")

num_17 ("seventeen")

However for Portuguese the caller would hear these system prompts:

daysofweek_Monday ("segunda-feira")

num_13 ("treze")

month_of_november ("de Novembro")

preposition_de ("de")

num_2 ("dois")

num_thousand ("mil")

conjunctive_and ("e")

num_17 ("dezessete")

Notice that audio for each language has the appropriate grammar, i.e. in en-US it uses "November" but in pt-BR it uses "de Novembro".

This feature is called runtime data playback. It's not supported for every language; it works in Portuguese (pt-BR) but not Belgium. When you add a supported language to a flow, Architect will put a * next to it if runtime data playback is not supported.

https://help.mypurecloud.com/articles/supported-languages-architect/

https://help.mypurecloud.com/articles/purecloud-supported-languages/

https://help.mypurecloud.com/articles/choose-supported-languages/

Hi Melissa,

I notice that “Portuguese”/“Portugal” has a *.

So this means that I cannot use the functions “ToAudioDate” ”, “ToAudioCurrency” or “ToAudioNumber”?

Even if I import my prompts as system prompts? As presented on the following image.

Thank you in advance.

Best regards,

Ana Laia

Hi Ana,

You can't add a generic language to a flow, you must specify the locale. So you can add Portuguese Brazil (pt-BR) or Portuguese Portugal (pt-PT) to a flow, but not generic Portuguese (pt).

When you go to import your own recordings to the system prompts, make sure you select the correct language (pt-PT not pt) or your recordings will not be heard in the flow.

Runtime data playback in not supported in Portuguese Portugal yet.