How to Use "Real-time speech to text" feature of Azure Cognitive Speech Service in ServiceNow?

Question

How to Use "Real-time speech to text" feature of Azure Cognitive Speech Service in ServiceNow?

If any of you has coded or know about this, please let me know

Answer 1

Hello Carmel,

There's no native ServiceNow plugin for Azure Speech-to-Text, you can implement it by using a custom UI widget + Azure WebSocket API or an external integration using MID server / integration hub / REST messages.

Custom UI Component with WebSocket Client

Use a Service Portal widget or Now Experience UI Framework to connect to Azure Speech.

Steps:

Create Azure Speech Resource

Go to Azure portal > Create Speech resource.

  Copy the key and region.
  
  **Enable WebSocket in browser (via JavaScript)**
  
     Use browser mic (`getUserMedia`)
     
        Stream audio via **Azure STT WebSocket endpoint**
        
           Parse and display the result in real-time
           
           **Embed this in a ServiceNow Widget or UI page**
           

GitHub sample JS client: https://github.com/Azure-Samples/cognitive-services-speech-sdk/tree/master/quickstart/javascript/browser

Option 2: Node.js Backend + ServiceNow Integration

If WebSocket handling in ServiceNow frontend is too complex:

Build a Node.js Express backend that:

Uses Azure Speech SDK (npm microsoft-cognitiveservices-speech-sdk)

  Accepts audio input (from browser or app)
  
     Returns transcript
     
     **Expose this backend as REST API**
     
     In **ServiceNow**, create:
     
        REST Message to call backend
        
           Scripted UI to handle audio upload and fetch transcription
           
              Use transcription in Incident/Case/Record creation
              

Option 3: Integration Hub + Azure REST API (Not Real-Time)

Use batch transcription with Integration Hub or REST Message.

Less real-time, but easier to set up.

You post audio blob → Azure returns transcript.

Azure Batch STT Docs

Best Regards,

Jerald Felix