Configuring voice input and screenshare in BMC HelixGPT

End users can use the  Voice input and Screenshare options to talk to AI agents and share their screen during support interactions in BMC Helix Digital Workplace.  These capabilities make it easier for users to explain issues and receive faster, more effective assistance. For example, when troubleshooting a ticket‑creation failure in BMC Helix ITSM – Incident Management, an end user can share their screen to show the exact validation error displayed during form submission and ask the AI agent for steps to resolve it.   

Supported audio and screen-share formats

• Audio input: PCM16, 16kHz.
• Audio output: 24kHz 
• Screenshare format: JPEG

Before you begin

• Allow WebSocket traffic through your firewall.
• Use the Voice Knowledge Tool if you want voice agents to retrieve information from OpenSearch DB.

Process to configure voice input and screenshare

To configure voice input and screenshare for your AI agent, the HelixGPT Agent Studio administrator and BMC Helix Digital Workplace administrator must perform the following tasks:

To create a skill with voice input and screenshare

1. Log in to HelixGPT Agent Studio.
2. On the Skills tab, from the Application list, select BMC Helix Digital Workplace > VOICE_EMPLOYEE_NAVIGATOR_SKILL.
3. Navigate to the Configure tab and copy the configuration.
   
   
4. Click Add skill and provide the following details:
   ​​​​​

   Field	Action
   Name	Specify a name for your skill.
   Type	Select the type of skill.
   Provider	Select the generative AI provider.
   Model	Select the GPT model for the selected provider.
   User context process definition	Select a process to evaluate the variables that you want to use in prompts.
   Instruction type	Select the instruction type.

    

5. Paste the copied configuration into the new skill and click Apply.
   
   
6. On the Agent tab, select VOICE_EMPLOYEE_NAVIGATOR_AGENT and click Apply.
   
   

To select a live model

1. In HelixGPT Agent Studio, open Settings.
2. Select HelixGPT > Connections > AI service providers.
   
   
3. Select the service provider and model used for your skill.
4. In Live model, select gemini-live-2.5-flash.​​​​​
5. Click Test Connection.

To map the skill in BMC Helix Digital Workplace

1. ​​​​​​Log in to the BMC Helix Digital Workplace admin console.
2. Select Configuration > Chat.
3. On the Chat panel, select the skill you created and click Save.
   
   

Result

Users can speak to the Employee Navigator AI agent, share their screen, and receive text or voice responses based on the selected live model.