Configure AudioCodes for your bot

  • Create a new, plain text secret for your engagement adapter. In this example, we’re giving it the name audiocodecsvoicesecret, and it’s value should be some unique secret string - a UUID is a good candidate here. Save both the created secret SRN, and the value of the secret - we’ll use both later.
  • Create a new endpoint for your bot of the following format:
    • (Replace references to yourorg with your organization name, voicebot with the name of your bot, and Secret with the SRN of the secret we just created)
  "Address": "yourorg-audiocodecsvoiceadapter",
  "OutboundBotReference": "voicebot",
  "InboundBotReference": "voicebot",
  "Organization": "yourorg",
  "EngagementType": "AudioCodesVoice",
  "Status": "online",
  "AllowedDomains": ["*"],
  "EngagementConfig": {
    "Secret": "srn:vault:eu-private-1:yourorg:secret:audiocodecsvoicesecret",
    "ContextParameters": [
  "Useragent": "useragent",
  "TargetBotReference": "voicebot",
  "Name": "yourorg voicebot adapter endpoint"

  • Create an AudioCodes bot of type AudioCodes Bot API. Give the bot some name.
  • Set up the Bot URL to be of the format https://conversation.$$ORG/$ENDPOINTADDRESS
    • Replace $REGION with your servisbot region (e.g. eu1, us1)
    • Replace $ORG with your organization name
    • Replace $ADDRESS with the address of the endpoint you created
    • Add any ContextParameters you may wish to collect - e.g. caller, callee - note this may collect PII in the form of a phone number. These properties will be available at root level in context, e.g. msg.payload.context.caller
  • Set the Token to be the value of the secret we created above.
  • Validate the bot configuration - it should result in a green success message. If not, check your configuration & try again.
  • Finish setting up audio codes - choose your STT and TTS vendors, along with the voice.

Test your Bot

Now, wire your AudioCodes bot to a purchased phone number, and try dialling it. You should be able to interact with your bot over voice.