Genesys Cloud - Developer Community!

Hello,

 nobody has any idea how could I achieve this?

Regards,

Mihai

------------------------------
Mihai Vasiloiu
Tech Lead Customer Interactions
------------------------------

I just implemented a very rudimentary version of this. I have a main in-queue flow that is used for hold music and messaging. It looks up the timings and the prompts to play based on the current queue that the interaction is in. It may play 3 prompts in a sequence with hold music in between each, then repeat the loop.

For example, it may be using a sequence like:

holdMusicTime = [30, 30, 60]

prompts = [VaccineMessage, MyTCMessage, ThankYouForHolding]

In that case it would play hold music for 30 seconds, then play the first prompt, play music for another 30 seconds, play the second prompt, ...

It uses a loop to iterate through those timings and prompts, so I have a step in the loop that does totalWaitTime = totalWaitTime + holdMusicTime[loopIndex]. I then have a Decision step that watches for totalWaitTime to hit certain thresholds so I can offer a callback, transfer to a different queue, etc. Since I'm not taking the length of the prompts into account, it's not exact. I don't really need it to be for my case, though.

Edit: I just reread your original message and see that you specifically called out the variability in prompt length which causes your timing to be inaccurate. In that case, I probably didn't help very much. :D I'm subscribed to this thread to see if someone else has a better idea, though.

------------------------------
Dave Halderman
Business Analyst
------------------------------


Original Message:
Sent: 09-02-2025 04:49
From: Mihai Vasiloiu
Subject: Time Based Events in Architect

Hello,

 nobody has any idea how could I achieve this?

Regards,

Mihai

------------------------------
Mihai Vasiloiu
Tech Lead Customer Interactions

As an idea, maybe you could get the current timestamp at the start of the in-queue flow, then calculate the timestamp (in the future) that you want the 'event' to occur, then each time you play a treatment (i.e. announcement or music) specify a timeout that equals the remaining time until the impending 'event'. When the treatment completes, check if you've reached the 'event' time, and if you have do whatever you want to do, but if you haven't then play the next treatment specifying a timeout that equals the remaining time until the impending 'event', and so on.

I just implemented a very rudimentary version of this. I have a main in-queue flow that is used for hold music and messaging. It looks up the timings and the prompts to play based on the current queue that the interaction is in. It may play 3 prompts in a sequence with hold music in between each, then repeat the loop.

For example, it may be using a sequence like:

holdMusicTime = [30, 30, 60]

prompts = [VaccineMessage, MyTCMessage, ThankYouForHolding]

In that case it would play hold music for 30 seconds, then play the first prompt, play music for another 30 seconds, play the second prompt, ...

It uses a loop to iterate through those timings and prompts, so I have a step in the loop that does totalWaitTime = totalWaitTime + holdMusicTime[loopIndex]. I then have a Decision step that watches for totalWaitTime to hit certain thresholds so I can offer a callback, transfer to a different queue, etc. Since I'm not taking the length of the prompts into account, it's not exact. I don't really need it to be for my case, though.

Edit: I just reread your original message and see that you specifically called out the variability in prompt length which causes your timing to be inaccurate. In that case, I probably didn't help very much. :D I'm subscribed to this thread to see if someone else has a better idea, though.

------------------------------
Dave Halderman
Business Analyst


Original Message:
Sent: 09-02-2025 04:49
From: Mihai Vasiloiu
Subject: Time Based Events in Architect

Hello,

 nobody has any idea how could I achieve this?

Regards,

Mihai

------------------------------
Mihai Vasiloiu
Tech Lead Customer Interactions

The limiting factor here is, no you can't interrupt a prompt that is already going. Only the built in Music on Hold prompt lets you set the amount of time to play.

So if you have a dynamic list of prompts for the music, at best you can know the duration of each prompt and from that make a decision before playing it, do I play just 1 prompt, or 2 or 3 before applying the treatment. You would have to store that prompt with it's duration in a a datable lookup, or api call, or store it in a list / json object but that could work. 

Or just take the average of all the prompt durations, and go on average customers will wait 2 minutes 30 seconds before they get their callback request for example. Sometimes less, sometimes more.