Skip to main content

Documentation Index

Fetch the complete documentation index at: https://semantiks.mintlify.app/llms.txt

Use this file to discover all available pages before exploring further.

The Conversations Panel is where you monitor and manage all active conversations between your AI agent and end users. It’s the primary workspace for human agents who handle escalations and provide live support.

​
Overview

The panel shows a list of conversations on the left, and the selected conversation’s details on the right. Each conversation includes the full message history—messages from the AI agent, the end user, and any human agent replies.

​
Conversation View

When you open a conversation, you see:
  • All messages in chronological order, including AI agent responses, user messages, and human agent replies.
  • Escalation indicators — visual pills that mark when a conversation was escalated to a human agent or unassigned back to the AI.
  • Timestamps on messages so you can track response times.

​
Human Agent Identity in Messages

When a human agent replies to a conversation, their messages display:
  • The agent’s name as it appears in their Semantiks team profile.
  • An avatar with initials next to their messages (visible to the end user in External Chat and Web Bubble channels).
This means end users can see who is helping them, making the support experience more personal and transparent. If the agent’s name is not available, a default label is shown.
The sender name is automatically included when you reply from the Conversations Panel. Make sure your team member profile has a name configured so it displays correctly to end users.

​
Message Grouping

Messages from the same sender are visually grouped together in the end-user chat interfaces (External Chat and Web Bubble). Consecutive messages appear with tighter spacing, and timestamps are shown only on the last message of each group for a cleaner reading experience.

​
Replying to Conversations

When a conversation is escalated:
  1. Select it from the conversation list.
  2. Type your reply in the message composer at the bottom.
  3. Send your message — it will appear to the end user with your name and initials.
Your reply is sent in real time. The end user sees it immediately in their active chat session.

​
Key Notes

  • Human agent replies are visually distinct from AI agent messages in the end-user interface (different colors and styling).
  • Escalation pills clearly mark transitions between AI and human agents.
  • You can review the full conversation history, including what the AI agent said before the escalation.