The Conversational Agents console is a new console that includes features from both the Dialogflow CX console and the Agent Builder console. All users of the Agent Builder console should start using the Conversational Agents console. Users of the Dialogflow CX console that want to incorporate generative features into agents should start using the Conversational Agents console.
Using the Conversational Agents console, you can create interactive agents that use generative playbooks, generative data stores, and deterministic flows.
The Dialogflow CX console will continue to be supported until the Conversational Agents console is generally available (GA).
Console compatibility
Agents previously built with the Dialogflow CX console or the Agent Builder console can be seamlessly accessed by the Conversational Agents console. Agents created with any of these three consoles use the same Dialogflow API.
Supported regions
This console supports all regions already supported by the Dialogflow CX console or the Agent Builder console.
Pricing
Pricing for agents created with the Conversational Agents console is the same as it has been for agents created with the Dialogflow CX console. Agents that utilize generative features such as playbooks, data stores, generators, or generative fallback for a request are charged as generative requests.
Create an agent
Prebuilt agents can be used to accelerate development for new agents in common scenarios. To create an agent with prebuilt agents:
- Open the Conversational Agents console.
- If you haven't already, select or create a project.
- Click Create agent.
- Click Use a prebuilt agent.
- Select a prebuilt agent and follow instructions for installation.
To create an agent with playbooks and/or flows:
- Open the Conversational Agents console.
- If you haven't already, select or create a project.
- Click Create agent.
- Click Build your own.
- Provide an agent name, location, time zone, and default language.
- Select a Conversation start option. If you want your agent to begin the conversation using a generative playbook, select Playbook. If you want your agent to begin the conversation using a deterministic flow, select Flow.
- Click Create.
- The console loads the editor page for the default playbook.
To create an agent with data store handlers:
- Open the Conversational Agents console.
- If you haven't already, select or create a project.
- Click Create agent.
- Click Create a Q&A Agent.
- You are redirected to the Agent Builder console. Follow the instructions to create an agent.
Top row navigation
The horizontal row at the top of the interface contains the following controls:
| X | Item |
|---|---|
| Project | The currently selected project. |
| Agent | The currently selected agent. |
| Language | The currently selected language used when entering language-specific data, and the Manage Languages option which brings you to flow language settings. |
| Search | Used to search, filter, and access core resources within the agent. |
| Collaborators | Collaborators that have access to the agent. |
| Export/Import agent | Agent export and import. |
| Change history | Agent change history. |
| Settings | Agent settings. |
| Publish agent | A convenience button for setting up Dialogflow CX Messenger. |
| Simulator | The simulator used for testing the agent. |
| Help | Links to documentation and other help content. |
| Feedback | Used to provide product feedback. |
Left navigation
| X | Item |
|---|---|
| Playbooks | List, create, and edit generative playbooks. |
| Flows | Visualize, create, and edit deterministic flows. |
| Tools | List, create, and edit playbook tools. |
| Prebuilts | Import ready to use prebuilt agents for a variety of common use cases. |
| Test Cases | List and define test cases to regularly verify agent behavior. |
| Conversation History | View the conversation history. |
| Environments | Create and edit various deployment environments. |
| Versions | Snapshot flow and playbook versions. |
| Integrations | Integrate your agent with ready to use user interfaces and other conversation platforms. |
| Conversation Profiles | List, create, and edit conversational profiles for advanced telephony use cases. |
Simulator
The simulator is used to test your agent. You can have a conversation with your agent while monitoring changes in session state.
Simulator settings
When you first open the simulator, you can configure the following settings:
- Agent environment
- Start resource, which can be a playbook or flow which receives the first input.
- The generative model (start resource is a playbook) or a page (start resource is a flow).
Simulator input
When interacting with the simulator, you provide end-user input as text, then press enter or click the send button.
In addition to plain text,
you can choose alternate input types by typing @,
then selecting the option:
- Channel: Send a channel value as a query parameter.
- DTMF: Send dual-tone multi-frequency signaling (Touch-Tone) input for telephony interactions.
- End user metadata:
Provide contextual hint in the form of a JSON object for generative AI to
better answer an end-user query.
For example,
for data store fulfillment or tool input, you could use:
{"firstName": "John", "lastName": "Doe"}. - Event: Invoke an event.
- Cloud Storage uri: Cloud Storage URI for a file to upload. You can use this file for images and documents that provide context to the generative AI.
- Parameter: Inject a parameter value. You can provide new parameters or provide preset values for existing parameters.
- Parameter scope: Set the parameter scope.
Simulator controls
Once the conversation has started, there are several controls available at the top of the simulator panel:
- Debug trace
- Original response
- Show latency
- Create test case
- Undo
- Replay conversation
- Reset conversation
- Options:
- Enable or disable webhook calls at any time with the webhook toggle button.
- Enable or disable partial responses.
- Enable or disable sentiment analysis.
- Save as example: See instructions
Simulator history
When using the simulator to have a conversation with your agent, the conversation history is shown in the simulator panel. You can click each conversational turn to expand more details about that turn. For agent responses, you can also preview the audio response by clicking the play button for the conversational turn.