Skip to main content

Omni Consultation Forms - Voice Agent

Updated this week

Online Consultation Forms allow your site to manage medical requests from your Patients via the Communications Window Forms Tab

The Voice Agent is designed to enhance accessibility by enabling patients to complete consultation forms over the phone. Once completed, the content of their call is captured and the response is added to the Communications window, ready for your staff to triage

Adding the Voice Agent to your Call Flow

To add and configure the Voice Agent, navigate to X-flow and click Load

Next select the call flow you wish to add this to, either by scrolling through the list or by using the Search box

Once your call flow has loaded, you can then begin to add the necessary blocks to facilitate the Voice Agent

First you will need to add in a prompt to inform your patients that they will be identified before interacting with the Voice Agent. Drag a prompt block from the left hand side to the workspace

Double click the Prompt Block to open up the editable properties. To set the Prompt audio, click the Audio tab

Use the Prompt Text box to configure the content your Patients when they call in. For Voice Agent usage, we recommend:

"We will now try to identify you, please speak clearly and concisely and try to ensure there is minimal background noise.

Press * at anytime to stop talking to our agent."

Click Save to close the Prompt box and keep your changes

For further guidance on configuring Prompt Blocks within your call flow, see our Help Guide here

Next, select the Assign Patient Block from the left hand side

And drag it onto the workspace next to your first prompt

Connect the Prompt and Assign Patient blocks by clicking and dragging the arrow to the connecting points

To edit the Name and Description, double click the Assign Patient Block and fill out the two fields. Click Save to make the changes

The Assign Patient Block has three outputs:

  • Identified

  • Not Identified

  • Default

Not Identified/Default Outputs

You will need to add a Prompt Block in to let patients know if they have not been identified

Drag another prompt block into the workspace next to the Assign Patient Block, and double click to open the properties. For Prompt Text, we recommend:

"I am sorry, we have not been able to identify you."

Click Save, and connect the prompt to the Not Identified and Default outputs by clicking and dragging the arrows from the connecting points

You will then need to ensure that the patient's call is then routed elsewhere in the call flow so that they can get assistance elsewhere

Identified Output

You will need to add a prompt in to let the patient know that they have been successfully identified, and to inform them that the Voice Agent will ask them questions about their call

Drag another Prompt Block into the workspace, and edit the text by double clicking the Prompt and entering it into the Prompt Text box. We recommend:

"Thank you, we have found your details. We will now ask you some questions about your call today.

Please speak clearly and concisely when answering questions and try to ensure there is no background noise. Our agent is limited to 10 minutes to capture your answers and is limited to 1 minute per answer.

We will repeat each answer back to you and ask you to confirm your response after each answer you give.

Press * at anytime to stop talking to our agent."

Once you have entered your text, click Save and connect the Prompt Block to the Identified output on the Assign Patient Block

Your Call Flow should now look like this

To Add the Agent Block to your Call Flow, click and drag it from the left hand side into the workspace, next to your Prompts

To edit the Agent Block's properties, double click the Block banner

Enter a Name and Description for the block by typing in the relevant fields

Set the agent to Medical Request Form

Click Save to close the Block and keep the changes

You will then need to add Prompt Blocks for the three outputs on the Agent Block:

  • Completed

  • Not Completed

  • Default

Completed Prompt

Drag a new prompt into the workspace next the Agent Block, and double click the header to configure the Prompt Properties. For Prompt Text, we recommend:

"Thanks for answering all of our questions, a member of the practice staff will get back to you regarding this problem."

Click Save to close the Prompt and keep the changes

Not Completed/Default Prompt

Drag a new prompt into the workspace next the Agent Block, and double click the header to configure the Prompt Properties. For Prompt Text, we recommend:

"I am sorry our voice agent hasn’t been able to assist you today."

Click Save to close the Prompt and keep the changes

Once you have added the Prompts, drag a Hang Up block into the workspace

Connect the Prompts to the Hang Up Block. Your call flow should now look like this:

To make the changes to your call flow live, click Deploy

Looking for more guidance on configuring your Call Flow within X-flow? Explore our Mastering X-Flow Academy Course here


Patient Experience

When a patient opts to use the Voice Agent to submit information, they will first need to be identified. They will be asked to state their Date of Birth for the in order to be verified

"Please state your Date of Birth"

Once they have stated this, the Voice Agent will repeat this back to them and ask the patient to confirm whether this is correct

"You stated your date of birth is [DOB], is that correct?"

The Date of Birth the patient provides is checked against the number they are calling from. If they are calling from a withheld number, they will be asked to call back

"We cannot detect the telephone number you are calling from, please call back ensuring your number is not withheld."

If they are not successfully validated, their call will be routed as per your call flow configuration - for example, to speak to a call handler directly

If their Date of Birth has been successfully validated, they will hear the prompt confirming this. They will then be routed through to the Agent Block, where they will follow a question and answer process to gather details relevant to their condition or request

They will be asked to answer the following:

  • "Is this regarding an existing problem? Please state Yes or No"

  • "Please describe the medical problem you're calling about today"

  • "How long has this problem been going on for? Is it getting better or worse?"

  • "Have you tried anything to help the problem?"

  • "Is there anything that is particularly a worry?"

  • "Tell us in a few words how we can help"

  • "Could you please tell us the best times to contact you"

Once the patient has completed their response for each question, the Voice Agent will repeat the information back to them, and ask them to confirm whether this is correct

If the patient confirms that the information is correct, the Voice Agent will move onto the next question

If the patient relays that the information captured is not correct, they will be asked to reconfirm their answer for the Voice Agent to recapture

If at any point the caller wants to leave the process of using the Voice Agent (either at the identification stage or during the questions and answers), they can do so by pressing the * key


Voice Agent Response

Once the Voice Agent has captured the response and the call has ended, the response is added as a new item in the Waiting Tab in the Forms area of the Phonebar Communications Window

From here, the response can be assessed in terms of Severity and Priority, and assigned to Users within your practice for triage

For more information on managing your Online Consultation Forms within the Phonebar, see our Help Guide here

If at any point the caller wants to leave the process of using the Voice Agent (either at the identification stage or during the questions and answers), they can do so by pressing the * key

When they do this, their partially completed response will be added to the Communications Window Forms area as an unfinished response in the Incomplete Tab


For further learning on the Voice Agent, explore our Academy Course here

Did this answer your question?