Conversations

Browse, search, and filter all conversations with search and filter tools

The Conversations page shows all conversations handled by your AI agents. Use the extensive filtering options to find specific conversations for review, analysis, or quality assurance.

Conversations List

Accessing Conversations

Navigate to Chat History in the sidebar.

You need Admin or Account Manager permissions to access the Conversations page.

Conversation Table

The table displays all conversations with the following information:

Column	Description
User	Email address of the user who started the conversation
Agent	The AI agent that handled the conversation
Messages	Number of messages exchanged
Created	When the conversation started
Evaluation	Manual evaluation status (OK/KO/Neutral)
CX Score	Automated quality score (if evaluated)
Topics & Tags	Applied topics and tags

Click any row to open the Conversation Management view where you can evaluate and manage the thread.

Primary Filters

These filters are always visible at the top of the page:

Filter	Description
Start Date	Show conversations from this date onwards
End Date	Show conversations up to this date
Agent	Filter by the AI agent that handled conversations
Ticket Created	Show only conversations that resulted in a ticket

Secondary Filters

Click “More filters” to access additional filtering options:

Filter	Description
Email	Search by user email address (partial match supported)
Evaluation	Filter by manual evaluation: OK, KO, Neutral, or Not Evaluated
CX Score Range	Filter by automated score: Poor (0-49), Needs Improvement (50-79), Good (80-100)
Search Content	Full-text search within conversation messages
User Type	Filter by user permission level
Organization	Filter by user’s organization
Topic	Filter by conversation topics (multi-select)
Tags	Filter by tags (multi-select)
Reviewer Email	Filter by assigned reviewer
Knowledge Source	Filter by knowledge source used in responses
Knowledge Page ID	Filter by specific documentation page referenced
Support Asked	Show conversations where human support was requested
Needs Review	Show conversations flagged for review by the LLM evaluator

Common Filter Combinations

Quality Review

• Needs Review = Yes
• CX Score Range = Poor or Needs Improvement

Escalation Analysis

• Support Asked = Yes
• Ticket Created = Yes

Agent Performance

• Agent = [Select agent]
• Evaluation = [Select status]
• Date range = Last 30 days

Topic Research

• Search Content = [keyword]
• Topic = [relevant topics]

Exporting Data

Click the Export button to download the current filtered results as a CSV file. The export includes:

• All visible columns
• Applied filters are reflected in the exported data
• Maximum 10,000 records per export

Apply filters before exporting to get a focused dataset for your analysis.

Navigation Features

Pagination

Results are displayed in pages of 50 conversations. Use the pagination controls at the bottom to navigate between pages.

Filter Persistence

Your filter settings are automatically saved in your browser. When you return to the page, your previous filters will still be applied.

Quick Navigation

After opening a conversation, use the navigation arrows in the conversation view to move to the next or previous conversation without returning to the list.

Best Practices

1. Start with date range: Narrow down by time period first, then add specific filters

2. Use “Needs Review” for QA: Focus your quality assurance efforts on flagged conversations

3. Combine filters strategically: Multiple filters work together (AND logic) to narrow results

4. Export regularly: Create periodic exports for trend analysis and reporting

5. Clear filters when done: Click “Clear filters” to reset and see all conversations

Related Features

• Conversation Management - Review and manage individual conversations
• Thread Evaluator - Configure automated evaluation
• Feedback Analytics - Analyze user feedback trends