Using Generative AI Tools Within Your CRM's Agent Builder

What Are Generative AI Tools?

Generative AI tools are specialized processing steps you can add to an agent flow within your CRM's agent builder. These tools enable the automated creation of text, audio, images, or videos in real time while the agent is active. Instead of relying on pre-written content, the agent can generate fresh, dynamic assets based on your instructions, stored brand guidelines, and data from the ongoing interaction.

Key Benefits

Integrating AI content generation into your agent workflows offers several advantages:

• Dynamic Content Creation: Produce tailored text, audio, images, and videos on the fly within an automation sequence.
• Content Reusability: Save the AI-generated output into a variable, allowing it to be referenced and used in later steps of the same agent flow.
• Brand Alignment: Apply your saved Brand Voice and Brand Kit settings to ensure generated content matches your company's tone and visual identity.
• Workflow Flexibility: Insert content generation at any point in an agent where dynamic asset creation is needed.
• Enhanced Personalization: Incorporate variables from previous steps (like contact details or user input) to create highly relevant and contextual outputs.

Adding a Generative AI Tool to an Agent

To insert a Generative AI tool into your agent's workflow, follow these steps:

1. Navigate to AI Agents from the main left-hand menu.
2. Open the Agent Studio tab.
3. Select the specific agent you wish to edit to open it in the builder canvas.
4. Click on the node in the flow where you want the content generation to occur.
5. Select Add a processing node.
6. In the panel that appears, find and open the Generative AI section.
7. Choose the type of tool you need: Text Generation, Audio Generation, Image Generation, or Video Generation.
8. A configuration panel will open on the side. Configure all required settings for the chosen tool.
9. Click Save to add the node to your agent flow.

Configuring the Text Generation Tool

The Text Generation tool is used to create written content such as responses, summaries, descriptions, or structured text.

Node Name

This required field identifies the step in your builder. A default name is provided, but you can rename it to something more descriptive, especially if your flow contains multiple processing steps (e.g., "Create Welcome Summary" or "Generate Product Recommendation").

Brand Voice

Use the dropdown menu to select a pre-configured Brand Voice that defines the communication style for the generated text. You can choose an existing voice, select None, or click + Add new to create a new one.

Text Prompt

This is the core instruction for the AI. In this required field, describe what you want the AI to create. Be specific about the task, any source information to use, and the desired format or tone. For example, instruct it to "Summarize the user's last message into three bullet points" or "Write a friendly follow-up email thanking them for their inquiry."

You can make your prompt dynamic by using the variable picker within the field to insert values from categories like Input Variables, Runtime Variables, Contact details, or the current date/time. Use the Enhance Prompt button for assistance in refining your instructions, and the expand icon to access a larger text area for complex prompts.

Variable Name

This field specifies where the AI-generated text will be stored. You define a variable name here (e.g., generatedSummary). The text output from this node will be saved to that variable, allowing you to use {{generatedSummary}} in subsequent steps of the agent, such as in a message or an email.

Configuring Audio, Image, and Video Generation

The configuration for Audio, Image, and Video Generation tools follows a similar pattern to Text Generation, with tool-specific prompt fields.

• Audio Generation: Provide a text prompt describing the audio to create (e.g., "A cheerful welcome message in a British accent"). The generated audio file URL will be stored in the variable you specify.
• Image Generation: Enter a detailed visual description in the prompt field. The generated image URL will be saved to your chosen variable.
• Video Generation: Use a detailed video prompt to guide the creation. The tool will generate the video output based on your configured setup and store it accordingly.

Using Variables with Generative AI Tools

Variables are powerful when combined with AI generation. You can pull data from earlier in the conversation (like a contact's name, a submitted answer, or a captured piece of information) and insert it into your AI prompt using the variable picker. This allows the generated content to be personalized and context-aware. The output of the AI tool is itself stored in a new variable, making it available for the rest of the agent's workflow.