How to setup Azure recognition

1. Log in to your Microsoft Azure portal

2. Check for the available azure subscription. If there is no subscription, you have to create a new one to use Azure services.

3. Create a new resource

4. Select Speech Services

5. Select available Azure subscription and resource group (if there is no resource groups, please create a new one)

6. Select your region to improve latency between your server and Azure Speech Recongition, Specify the name for the resource and select pricing tier.

Free pricing tier could be used for evaluation purposes.

Paid tier prices depend on the region you've selected.

7. Select networks available for the recognition service

If your 3CX server is hosted outside of the Azure cloud, you should use option "All networks".

8. You could skip the Idenity step if there is no your personal needs for set up

9. Tags step is also not needed for this installation

10. After the speech service is created, open it up

11. Find the asr.config file in the webapi root directory (/var/www/webapi for linux, C:\Program Files\3CX Phone System\Bin\webapi for windows)

Copy the KEY 1 from speech service page

Replace the YOUR_AZURE_SUBSCRIPTION_KEY value in the asr.config file, azure row

12. Copy the region name from endpoint address on the speech service page

Paste it replacing the YOUR_AZURE_REGION value in the asr.config file.

Region value should look like "northerneurope", "eastus", "uksouth".

13. Save asr.config

14. Restart 3CX WebAPI service

• Windows

  • sc stop nginx 

  • sc start nginx

• Debian

  • sudo systemctl restart webapi

15. Setup is finished, further call transcription will be provided by Microsoft Azure Speech Service. 

Integration log will be stored in asr.log file