Skip to main content
The conversations page at /platform/{chatbotId}/conversations provides a browsable, filterable table of every conversation your chatbot has had. Click into any conversation to see the full message thread with AI analysis annotations.

Conversation list

The table displays the following columns:
ColumnDescription
UserName of the end user
EmailUser’s email address
OrganizationUser’s organization
MessagesMessage count in the conversation
SentimentAverage sentiment with color-coded badge
PlanUser’s subscription plan
Last messageTimestamp of the most recent message
CreatedWhen the conversation started
Click any column header to sort the table by that column. Click again to reverse the sort order.

Search and filter

Use the search box to filter conversations by keyword. For advanced filtering, click the Filter button to open the filter popover:
Filter popover with fields for User Industry, Org Plan, Device, Language, and more
Available filter fields include:
  • User Industry
  • Org Plan
  • Device
  • Language
  • Template name
  • Messages (count range)
  • Sentiment average (score range)
  • User MRR
  • Org MRR

Column visibility

Click the View button to show or hide columns. There are 11 toggleable columns.
Column visibility toggle showing 11 toggleable columns

Date range and pagination

Use the date range button to filter conversations by time period (defaults to This month). Pagination controls at the bottom let you set rows per page (10, 20, 50, or 100) and navigate between pages.
Pagination controls with rows per page selector

Conversation detail

Click on any conversation row to open the full thread at /platform/{chatbotId}/conversations/{conversationId}.

Message thread

Each message shows the sender (user or assistant) and the message text. AI analysis annotations are displayed as clickable highlights and tags:
  • Sentiment tags — color-coded sentiment scores (e.g., “Very Negative”, “Positive”)
  • Behavior flags — detected user behaviors (e.g., “Frustrated User”, “Buying Signal”)
  • Quality flags — assistant quality issues (e.g., “premature closure”, “hallucination”)
  • Resolution status — how the conversation was resolved (e.g., “Fully Resolved”)
Click on any annotation to see the AI’s reasoning and detailed analysis:
Sentiment annotation detail panel showing score, reasoning, and message preview
The right sidebar shows contextual information about the conversation:
Aggregate sentiment score and distribution across all messages in the conversation.
Message count, conversation ID, start and end timestamps.
Name, email, user ID, subscription plan, industry, and MRR (monthly recurring revenue).
Organization name, org ID, plan, industry, and MRR.
Session ID, page URL, referrer, device type, and country.

Next steps

Deep search

Find conversations by meaning instead of keywords.

Analysis config

Customize which annotations OpenBat generates.

Client management

Track organizations and users across conversations.