Reading transcripts

What every element in an open conversation shows — header, summary banner, message bubbles, source citations, timestamps.

Once you pick a conversation from the list (see Conversations overview), the right panel shows the transcript. This page walks through everything you'll see there.

The header has a three-row info block on the left plus status action buttons on the right.

Left block:

  • A circular avatar showing the first letter of the visitor's name (or ? for anonymous)
  • Visitor name — the visitor's name if known, otherwise a short identifier for their session
  • Visitor email — only shown if known. Clickable to open your email client.
  • Visitor phone — only shown if both email and phone are known. Clickable to call.
  • A row of metadata: a channel dot (green for WhatsApp, blue for Telegram, muted for widget), "{N} messages", and a status badge.

Status badges:

  • Needs reply (amber dot + pill) while the conversation is waiting for a human
  • Closed (muted pill) once the conversation is closed
  • No badge for active conversations

Right block — status action buttons (see Replying as an agent for what they do).

On small screens, a left-pointing chevron sits in the leftmost position of the header to take you back to the conversation list.

Handoff summary banner

A banner appears when a conversation is waiting for a human and the AI has already written a short summary for you.

┌──────────────────────────────────────────────────────────┐
│  Summary:                              │
└──────────────────────────────────────────────────────────┘

It shows inside the panel, below the header, with an amber border and background. Format: a bold Summary: prefix, then the summary itself.

How the summary is generated

When the chatbot hands off to a human, a background job writes a short summary of the conversation so the teammate picking it up can catch up quickly:

  • Only the last 20 messages from the conversation are used.

  • The prompt behind the scenes is:

    Summarize this customer support conversation in 2-3 concise sentences. Focus on: (1) what the user is asking for, (2) what the AI has already tried, (3) why a human agent is needed. Do not add pleasantries — just the summary.

  • If the model returns something useful, it's saved on the conversation and the banner appears. If generation fails for any reason, the banner simply doesn't show — the rest of the transcript is unaffected.

The summary takes a few seconds to generate. Until it's ready, the banner is absent. Once it's ready, the banner appears automatically — no refresh needed.

Message thread

Each message shows as a bubble. Layout depends on who sent it:

  • Visitor messages — right-aligned, primary color
  • Agent (human) messages — left-aligned, blue-tinted background, with the agent's name above the bubble
  • Bot messages — left-aligned, muted background

Each bubble contains:

  1. The agent's name (only on agent messages)
  2. The message content — line breaks and leading spaces are preserved
  3. Sources footer — appears only when the bot's reply cites sources. Format: "Sources: [list]".
  4. Timestamp in the bottom-right corner — HH:MM in your local time.

Reply box

Only shown once a conversation is waiting for a human. See Replying as an agent.

Auto-scroll

When a conversation opens and whenever a new message arrives, the thread scrolls to the bottom automatically. There's no manual scroll-to-bottom button — new messages just appear at the bottom.

What the header does NOT show

  • No agent assignment or team handoff — any logged-in member can reply
  • No "first responded at" metric
  • No chatbot indicator (you're already inside a chatbot's conversations tab)
← Previous
Conversations overview
Next →
Replying as an agent