Google Cloud dialog agent constructor

Dialogflow CX.

Generative dialogue with explicit transitions between stages

4.65 / 5Directory score
Official website

Dialogflow CX accepts text or voice queries. Agent behavior is described through threads or generative playbooks.

Google Cloud documentation on Dialogflow CX capabilities
Dialogflow CX documentation about streams, scripts and supported channels. Open the source

Choose where the dialog requires control

Customers ask the same question in different words. For free clarification, a generative scenario with instructions and examples is convenient. If the process requires a strict sequence, break it down into obvious stages. Dialogflow CX documents both approaches. Before setting up, write out where the agent should clarify the data, and where he can only show a help or transfer the appeal further. It is easier to check that the conversation does not jump through the mandatory stage.

Training request for change of delivery

In a fictional scenario, a customer names a city and corrects it in the next turn. Check which city the agent stores and uses in its response. Test a separate case where the required field remains empty. If an internal system has to fulfill the request, test that branch with a test integration. Confirm the update using the connected service's status, since the model response alone cannot establish that it was saved.

Repeat checking for text and voice

Run the same scenario in the channels you plan to use. Voice input can mishear a street or number even when a text test passes. A confirmation turn before the action can help. Test what happens when the external system fails and how the conversation ends. Decide what the customer will hear or see if the request cannot be fulfilled now. Record expected transitions for retesting after changes and document which fields the integration validates.

Suitable tasks

  • Gather an assistant for typical calls.
  • Check that the user has fixed an important field in the dialog.
Features

Strengths

  • Supports explicit transition management and generative scenarios.
  • Accepts text and voice input.
Things to consider

Limitations

  • External actions require integrations and rights.
  • The script should be checked for each channel separately.

Selection and conditions

Free access
Not confirmed
Trial period
Not confirmed
No sign-up required
Not confirmed
Russian interface
Not confirmed
Local installation
Not confirmed
API
Yes

“Not confirmed” means there is not enough information. The feature may still be available; check the developer’s terms.

Input formats
Text, Audio
Result
Audio, Text

Terms

For the test, use a separate project and test channel.

From task to first result

Practical example

Task
Gather an assistant for typical calls.
What to prepare
Training number and agreed certificate of a fictional workshop.

Task brief or prompt

Set up an agent to respond about work hours and address. In the test call, interrupt the answer with a question about the address, then ask to transfer the call to an employee.

Expected result

Test conversation with a verifiable answer and transition to an employee.

How to check the result

  • Check the opening hours and address with a certificate.
  • Repeat the translation when the employee is unavailable.

An editorial teaching example of a task brief, not the result of hands-on testing of the service. Check that the required mode is available, that you have rights to the source material and that the intended use of the result is permitted.

Check details

Official sources

  1. Dialogflow CX documentation
  2. Guide to generative scenarios
  3. Tariffs of Conversational Agents

This description is based on the developer’s material. Features, prices and availability may change; check the primary source before use.