Panel Sections
Agent Information
At the top of the panel, you will see the Chat Agent that handled this conversation:- Agent name with a link to the agent’s configuration page
- Channel icon showing the communication channel (website, SMS, WhatsApp, etc.)
- Agent description providing context about the agent’s purpose
- Agent version if versioning is enabled (e.g., “v2” or “Draft”)
Contact Information
For known contacts (visitors who have shared their details), this section displays:| Field | Description |
|---|---|
| Name | The contact’s name, linked to their full contact profile. Click to open in a new tab. |
| The contact’s email address (click to copy) | |
| Phone | The contact’s phone number (click to copy) |
Boot Data
If the Chat Agent was configured with boot data (custom data passed when the chat widget loads), this section shows the key-value pairs. This is useful for understanding the context the visitor had when starting the conversation. You can toggle between a formatted view and a raw data view.Root URL
If the conversation started on a website, this section shows the page URL where the visitor initiated the chat. Click the link to visit that page.Metadata
If the conversation has global or thread-level metadata attached, you can expand this section to view it. Metadata is organized into:- Global metadata — shared across all conversations for the contact
- Thread metadata — specific to this conversation
Analysis and Summary
This is one of the most valuable sections. When your Chat Agent has analysis configured, this section shows:- Summary — an AI-generated summary of the conversation, rendered in formatted text with support for headings, lists, and links
- Analysis Results — structured data extracted by the AI based on your analysis configuration (e.g., conversation category, lead qualification score, extracted fields)
Visitor Feedback
If the visitor left a rating at the end of the conversation, this section displays:- Star rating (1-5)
- Written feedback message (if provided)
- Timestamp of when the feedback was submitted
Tickets
The Tickets section lets you create and manage support tickets linked to the conversation:- Click New to create a ticket directly from the conversation
- Existing tickets are displayed with their status
- Click the expand arrow to collapse or expand the section
AI Evaluation
This section provides quality control for your AI agent’s performance:- Mark as failed — toggle this switch if the AI agent did not handle the conversation correctly
- Use this to track and improve your agent’s accuracy over time
Comments
At the bottom of the panel, the Comments section is where your team can leave internal notes. See Managing Conversations for details on how to use comments.Tips
- Expand the Analysis and Summary section after reviewing a conversation to quickly understand what happened without reading every message.
- Click the contact name to jump to their full contact profile, where you can see all their conversations and contact history.
- Use the Agent link to quickly navigate to the Chat Agent’s configuration if you need to adjust its behavior based on conversation outcomes.
- Mark conversations as failed when the AI makes mistakes — this data helps you identify patterns and improve your agent’s instructions.
- Check the Root URL to understand the context of website conversations — knowing which page the visitor was on can explain their questions.