Talkdesk Express: FAQs - Opt-out of AI capabilities

 

Can I disable the Out-of-the-box configured AI capabilities?

Yes, it is possible to disable the AI capabilities that are offered from the start in just a few minutes, if you wish to not use them, or if you want to explore different configuration.

At start, and out-of-the-box, a series of AI capabilties are enabled by default:

  • Copilot's Summarization and Disposition
  • Copilot Flows
  • Rewriter
  • Translator
  • Knowledge Management Source to AI tools
  • Identity - Phone Validation
  • Identity - Voice Biometrics
  • Navigator
  • Text-to-Speech (TTS) with premium voices

This document will guide you trough the process of deactivating the active configuration of these capabilities on the account. 

Note: The deactivation can be partial to only disable some of the features, keeping others enabled.

 

Steps to disable AI capabilities

Step 1

Navigate on the left menu to the Admin section and, then, select AI Launchpad.


1 - Open AI Launchpad.png

Step 2

On AI Launchpad, click on Copilot's Summary feature and then, on the right side bar, deactivate all the available options by switching the toggle from ON to OFF and click on "Save options".

2- Copilot Off.png

Step 3

Navigate to the Default AI Flow section and click on the menu item on the right. Then, swith the toggle from Active to Inactive.

3- Deactivate Flow Copilot.png

Step 4

Still on AI Launchpad, scroll down to the Conversations section.

Here, select the "Rewriter" option and, on the right side panel, switch the toggle from On to Off and click on "Save options".

4 - Deactivate Rewriter.png

Step 5

At this step, on the same Conversations section of the AI Launchpad, select the "Translator" option and, on the right side panel, switch the toggle from On to Off and click on "Save options".

5 - Deactivate Translator.png

Step 6

On the left side Menu, navigate to Knowledge Management app and, then, select the "External Sources" option.

On this list, open the 3 dots menu on the right for the desired source and click on Deactivate.

6 - Deativate KB.png

Step 7

From this step foward, you will be be removing the AI capabilities from Studio Flows, which means that the pre-build flow to handle Voice and Digital interactions will need to be adjusted to guarantee that the removal of the AI components do not leave an incoherent flow.

There are two flows pre-configured on the account, one for Voice and one for Digital interactions, that will require adjustment.

To achive such, navigate to Studio on the left side menu. From this screen, to edit a flow, please click on the "View Published" button of the desired flow to edit.

7 - Enter Studio Flows.png

Step 8

Once a Studio flow is opened, a the flow will appear as "view only". To edit, click on the "Edit" button on the top right corner.

8 - Edit Studio Flow.png

Step 9

Starting by the "Inbound Digital Flow Template", for digital interactions, three components need to be removed:

  • Analyze digital message - Navigator Digital
  • Validate Phone - Identity SMS Phone Validation
  • Validate Phone - Identity WhatsApp Phone Validation

To do so, when each component is selected, click on the red garbage bin on the left of the component.

Perform this action for the three components.

 

9 - Remove AI Components from Digital.png

Step 10

With the previous step completed, the Studio Flow is now incomplete and requires adjustments to compensate for the removed options.

Select the first component "Incoming Message - Initial Step" and, on the right side panel, select the "Exit" tab. Here, on the "Ok" option, choose the "Channel Identification" component.

10 - Change Initial Step to Conditional Statement.png

Step 11

On this step, open the components "Automated Message - SMS Reply" and "Automated Message - WhatsApp Reply" and for both, on the right side panel, select the "Assign to Agent" option on the "OK" Exit.

 

11 - Change SMS and Whatsapp Reply to Assign to Agent.png

Step 12

At this stage, the flow is now completed and functional. Please, click on "Save" and then "Publish" on the top right corner.

12 - Save and Publish Digital.png

Step 13

Please, open the "Inbound Voice Flow Template" as mentioned on Steps 7 and 8 to disable the AI capabilities.

On this flow, five components need to be removed:

  • Validate phone - Identity Phone Validation
  • Analyze voice message - Talkdesk Navigator no EWT
  • Analyze voice message - Talkdesk Navigator Less than 5m
  • Analyze voice message - Talkdesk Navigator more than 5m
  • Authenticate voice - Identity Voice Biometrics

To do so, when each component is selected, click on the red garbage bin on the left of the component.

Perform this action for the three components.

13- Remove AI components from Voice Flow.png

Step 14

With the previous step completed, the Studio Flow is now incomplete and requires adjustments to compensate for the removed options.

Select the first component "Record - Record Call" and, on the righ side panel, choose the "Estimated Waiting Time" option on the "Ok" exit.

14 - Add exit on Record Call.png

Step 15

Now, select the "Estimated waiting time - Estimated Waiting Time" component and, on the right side panel, select the "Estimated Waiting Time Options" option on the "Successful exit" of the "Exits" section.

15 - Add exit on estimated waiting time.png

Step 16

Select the "Conditional Statement - Estimated Waiting Time Options" component and, on the right side panel, let's add a new step on the "Condition #3" on the "Exits" section.

On this "Condition #3", go to the "Sent to" option and select "Add new step".

16 - Add new IVR Step.png

Step 17

On this new Step, choose the "Voice IRV" component type and, for the step name, input "More than 5 min". Click on Add step to confirm.

17 - Add Voice IVR.png

Step 18

On this step, select the newly created "Voice IVR - More than 5 min" component.

On the "Preferences" tab on the right side panel, input the following "Message to output":

"Sorry, but our estimated waiting time to answer your call is more than five minutes. Would you like us to call you back later to this same number?"

18 - IVR Config.png

Step 19

Now, enter the "Exits" tab on the righ side panel and, on the "Voice Command #1", input the following keywords:

  • yes
  • si
  • sure
  • callback
  • call me
  • yup

Also, input "Callback" on the "Exit Name" and choose the "Callback" option on the "Send To" section.

19 - IVR Exit 1.png

Step 20

Scrolling down on the "Exits" tab, select the "Voicemail" option on both the "No match" and "Timout" sections.

20 - IVR Exit 2.png

Step 21

With the new "Voice IVR" component completed, now open the "Conditional Statement - Estimated Waiting Time Options" component and, on the "Exits" tab of the right side panel, choose the "Assign to an available Agent" option on the "No Match" section.

21 New - Add no match to estimated waiting time options.png

Step 22

At this stage, the flow is now completed and functional. Please, click on "Save" and then "Publish" on the top right corner.

22 - Save and publish voice.png

 

Step 23

In the case here is the need to disable Premium Text-to-Speech (TTS) and it's associated costs, the setting change is simple to be achieved. Existing Text-to-Speech settings on components within the "Inbound Voice Flow Template" Studio Flow are configured to use "English, United States" as the selected launguage.

By default, the voice configured for that Language is a premium one, which has associated costs per use.

To change this configuration, please go to Admin -> Channels -> Voice, as represented on the next image.

23 - remove TTS premium.png

 

Step 24

Inside the Voice option, please navigate to the "Text to Speech voices" tab and search for the "English, United States", language. You can also use the search box to find the language quickly. 

After locating the line, click on the edit button on the right side (pencil logo).

 

24 - remove tts premium 2.png

Step 25

Now, inside this language, you can see several available voices that can be used on the "English, United States" language. For each, on the right, you can see if it is a Standard, Advanced or Premium voice. Standard voices are free of charge, while Advanced and premium are charged per use.

The voice existing out-of-the-box is a Premium voice so, to deactivate it, please select a "Standard" voice type, as portraited in the image bellow.

Once selected, click on the "Change" button on the top right so save the configuration.

 

25 - remove tts premium 3.png
All Articles ""
Please sign in to submit a request.