Google Dialogflow

Last Updated : Jun 05, 2026 |

Avaya Experience Portal supports full native integration of Google Dialogflow. Using Google Dialogflow, customers can access Cloud AI based automation for voice calls using Experience Portal.

Dialogflow speech recognition extracts intents from customer conversations. Dialogflow then uses these intents to automate processes such as room booking, reservations, and find relevant answers from FAQs. For more information on how these bots are developed on the Google Cloud platform, see https://dialogflow.com/.

Avaya Experience Portal provides the telephony gateway and call management functionalities to compliment Dialogflow. Avaya Experience Portal streams audio from a caller to Google Dialogflow using the open source gRPC and acts on responses from Dialogflow. For example, transfer a call, collect DTMF, or play audio file/DTMF.

Experience Portal provides out-of-the-box integration with Google Dialogflow for voice applications through a default VXML application on MPP. This default VXML application on MPP acts as the main interface for integrating Avaya Experience Portal with any Dialogflow bot. No changes are needed to the default Dialogflow application on MPP.

Avaya Experience Portal Dialogflow integration supports the following capabilities:

  • Audio streaming between Avaya Experience Portal and Dialogflow

  • VXML transfer (bridge, blind, and consultative transfers) and notification of transfer complete or failure (bridge transfers)

  • DTMF transfer using feature access codes, such as playing DTMF into call

  • DTMF dial pad collection by MPP using TELEPHONY_DTMF event

  • SIP Header and avaya-session-telephone session parameters using contexts in Welcome event

  • VXML privacy support

  • VXML timer support

    • Experience Portal can set no_input_timeout (VXML property: timeout) and speech_complete_timeout (VXML Property: completetimeout).

  • Controlled number of attempts

    • MPP sends NO_INPUT when VXML silence is detected. This is used by the bot to control the number of attempts.

  • Fetch audio support

    • Playing audio files for an unspecified amount of time until the follow up event is received from Dialogflow.

  • Invalid prompt (No match)

  • Barge-in control

  • Playing of pre-recorded prompts

  • Sending call disconnect event (Hangup) to Dialogflow when the caller hangs up