DialogFlow ES virtual agent flow template

Overview

This flow showcases how data is passed between Webex Contact Center and DialogFlow ES for processing customer interactions. It provides a foundational flow where data is exchanged with DialogFlow ES for natural language processing and automated agent fulfillment. The integration with DialogFlow enables the bot to understand customer intents and take appropriate actions based on the conversation. Additionally, the flow includes error handling to ensure a smooth customer experience, even when unexpected conditions arise.

Prerequisites

Ensure that the following requirements are met in the Webex Contact Center Management Portal before implementing this flow:

A Google DialogFlow ES agent with relevant intents for the conversation.
Create entry points, queues, teams, and entry point mappings.
Enable Webhook Fulfillment in DialogFlow ES and use the sample node.js code in the inline editor.
Cisco Text-to-Speech (TTS) is enabled for generating custom messages dynamically. Upload static audio files if custom audio prompts are used instead of Cisco Text-to-Speech (TTS).

Flow breakdown

The following table describes the various flow elements involved in the call process, detailing the actions and responses that occur during each stage.

Flow element	Description
Customer initiates contact	The call is received by Webex Contact Center.
Data is passed to DialogFlow ES	A custom greeting, which includes customer details such as name and reason for calling, is sent to the DialogFlow ES bot for processing.
Bot interaction with DialogFlow	DialogFlow processes the input and responds based on configured intents.
Music in queue	While the bot processes the request, the customer is placed in a queue with hold music.
Disconnect	The interaction ends once the dialog is completed.

Flow activities

The following table describes the sequence of flow activities involved in this flow template.

Flow activity	Description
Start	This activity marks the beginning of the flow. It's triggered when a new call is received.
Set Language	The flow uses a Set Variable activity to configure the language code (en-US) for the entire interaction. This ensures all voice interactions align with the caller's language preference.
Custom Greeting	This activity passes customer details such as name, email, and reason for calling to the DialogFlow ES bot. The greeting is dynamically generated using Cisco Text-to-Speech (TTS). Example data passed: customerName: Jane Doe customerEmail: customer@email.com customerReason: Bookings
Queue to Agent	If the interaction requires escalation, the customer is placed in a queue, and hold music is played using the `defaultmusic_on_hold.wav` file.
Play Music	Music is played while the caller waits in the queue. The flow uses Cisco’s default hold music but can be customized by uploading different music files.
Disconnect	This activity disconnects the call once the flow is complete, ensuring a seamless end to the interaction.

Flow specifics

The flow JSON used in this example contains variables and activities essential for interaction handling, error processing, and communication between Webex Contact Center and DialogFlow. The key variables used include:

Flow variable	Description
Global_FeedbackSurveyOptIn	Tracks whether the customer opts in for a post-call survey.
customerName	Captures the customer's name for personalization.
customerEmail	Captures the customer’s email.
customerReason	Records the reason for the customer's call.
Global_Language	Configures the default language (en-US).
Global_VoiceName	Determines the voice used for text-to-speech.