When creating or editing an Agent, three tabs are shown. The first tab, titled Configuration, is divided into three main panels:
- Agent Details
- Agent Setup
- Agent Presentation
This section allows you to define the core identity of your Agent. It includes the following fields:
This field is used to assign a clear and recognizable name to your Agent. This name identifies the Agent throughout the platform. (Required)
Here you define the main purpose of the Agent—what task it solves or what assistance it provides to the consumer. This description guides the Agent's behavior and provides the context needed to respond accurately. For example, if the Agent is intended to provide flight cancellation support, that should be its purpose. (Required)
This field complements the Purpose by describing the Agent's role in a short phrase, similar to a job title. While the Purpose indicates what the Agent does, the Role reflects who the Agent is or how it presents itself. For example, if the purpose is to help with cancellations, the role might be "Airline Customer Service Assistant".
Based on the title and description provided, The Lab can automatically generate an avatar. Click on Generate image to open a modal with different avatar options. You can choose one or click on Regenerate to generate a new set. Once satisfied, click on Accept to assign the selected avatar to your Agent.
This section defines the key elements that determine how the Agent should behave during interactions. It includes three main fields:
Here you can include contextual information or essential knowledge that the Agent must be aware of to respond correctly. This can include product descriptions, internal company processes, or other relevant data. This information helps the model generate more accurate answers aligned with the Agent's specific domain.
In this field, you define clear instructions on how the Agent should behave in different scenarios. For example, expected tone of voice, mandatory steps before replying, or policies to follow. Guidelines ensure consistent responses and help align Agent behavior with project objectives.
Note: You must complete at least one of the two fields (Background Knowledge or Guidelines) as they are essential for defining what the Agent should do and how it should respond. Both fields are incorporated into the prompt structure that guides the Agent’s behavior. For more details about how this structure is built, see
Agent Prompt composition.
This section allows you to define input/output data pairs that illustrate how the Agent should respond to specific messages. To add a new example, click on + Add new example. This will open a row with two columns: Input Data (for the input message) and Output Data (for the expected response from the Agent).
To delete an example, click on the Delete icon. This will activate the Confirm and Cancel icons, which allow you to confirm or cancel the deletion, respectively.
This section defines how the Agent is introduced to consumers in the Workspace. It allows you to configure a welcome message, short description, sample conversation starters, and a list of key features. These elements clarify the interaction and guide the consumer from the beginning. It includes four main fields:
This field is used to display a short message at the start of the interaction. The introduction can contain up to 100 characters and is intended to greet the consumer and set the tone for the conversation.
In the Workspace, the introduction appears as the highlighted welcome text at the top of the Agent Presentation. If left blank, the following default text is displayed: “Hello! I am an AI-Driven assistant”.
Here you can provide additional context about what the Agent can do and how it can assist the consumer. The description supports up to 200 characters and helps clarify the scope and purpose of the Agent.
In the Workspace, the description appears just below the introduction, giving consumers a concise explanation of the Agent’s role. If left blank, the following default text is displayed: “Whether you need to briefs, craft talking points, or simply seek information, I'm here to help!”
This option allows you to define predefined questions that appear when the chat begins. These questions serve as examples that help the consumer understand what kind of queries they can make. You can add up to five conversation starters, making the interaction more intuitive and reducing hesitation at the beginning.
To add a new conversation starter, click on + Add Conversation Starters. This opens a window where you can enter the Starter Title and the Starter Prompt.
- The Starter Title is a short label that summarizes the idea of the question and is displayed in the Workspace as a clickable option.
- The Starter Prompt is the complete question that is placed in the message input field when the consumer clicks on the title, so it can be reviewed, edited if needed, and then sent.
After entering the title and the prompt, click on Save to confirm. Once saved, the conversation starter will appear in the list and will be displayed to the consumer at the beginning of the interaction.
In the Workspace, the consumer sees the Starter Title (for example, Starter Title 1). When they click on it, the Starter Prompt (for example, Starter Prompt 1) is automatically placed in the message input field.
This option allows you to highlight the main qualities or distinctive aspects of the Agent. Features emphasize what makes the Agent useful, secure, or efficient—such as response accuracy, privacy compliance, or domain-specific knowledge. You can add up to three features, which will be displayed in the Workspace as cards below the introduction and description.
To add a new feature, click on + Add Features. This opens a window where you can enter the Feature Title and the Feature Description.
- The Feature Title is a short label that identifies the capability or quality you want to highlight.
- The Feature Description provides a short explanation—up to 200 characters—with more detail about the selected capability.
After entering the title and description, click on Save to confirm. Once saved, the feature will appear in the list and will be displayed in the Workspace as a card showing the title and description together.
In the Workspace, the consumer sees each feature as a visual block under the introduction and description (for example, Feature Title 1 with Feature Description 1). These features serve as quick references that reinforce the Agent's key strengths and help guide the consumer’s expectations regarding its capabilities.
