GitHubContribute in GitHub: Edit online

copyright: years: 2017, 2023 lastupdated: "2023-01-03"


Advanced configuration

After you have a basic setup, you can further configure components in the IBM® Voice Gateway architecture to provide additional capabilities.

New to IBM Voice Gateway? See the Getting started section to quickly get up and running with a self-service agent or agent assistant.

For self-service agents

  • Set up a SIP trunk - If you tested your setup with a simple SIP client such as Linphone, set up a SIP trunk as part of your final architecture.
  • Define your configuration - Using Docker environment variables, you can enable additional capabilities and configure the voice gateway's settings.
  • Program your agent using the API - The API uses action tags and state variables to exchange information between the IBM Watson™ Assistant service and your voice gateway. With the API, you can define when to hang up a call, set the music on hold to your organization's favorite smooth jazz track, or even dynamically configure the Watson speech services.
  • Set up a multi-tenant JSON configuration - Host multiple phone numbers from a single Voice Gateway instance for simpler, centralized administration.
  • Integrate with an IVR or contact center - To cover all your bases, you can give customers a chance to transfer to a live agent by integrating with an existing contact center.
  • Configuring call transfer - Using the API and configuration variables, for example, you can trigger a call transfer if a Watson service is down.
  • Configure call audio recording - Capture audio from customer callers and the Watson Text to Speech service in downloadable WAV files.
  • Configure real-time transcription - Automatically keep a written record of each conversation by setting up real-time transcription. Use these logs to monitor conversations or to analyze them to fine-tune your Watson services.
  • Generate reporting events for analytics - You can configure the voice gateway to generate events for call detail records (CDRs), Watson Assistant turns, and utterance transcriptions, then review and analyze the generated JSON data.

For agent assistants

Agent assistants take a little more work to initially set up, so less configuration is needed later.