Skip to content

Chat Conversations

Manage Live Chat conversations from the web widget.

Viewing Conversations

Conversations Page

Navigate to Conversations page to view all chat conversations:

  • Channel Filter: Filter by channel (Voice, SMS, Chat, Email, or All)
  • Search: Search by contact name, email, phone, or company
  • Date Range: Filter by conversation start date
  • Qualification Status: Filter by qualified/unqualified
  • Real-Time Updates: Conversations update automatically

Conversation Details

Click View Details to see:

  • Full conversation transcript
  • Contact information
  • Conversation metadata
  • AI conversation guide
  • Notes and tags

Chat-Specific Features

Contact Information Collection

Required Before First Message:

  • Name: Customer's full name (required)
  • Email: Email address (required)

Collection Process:

  1. Customer opens chat widget
  2. Widget displays information collection form
  3. Customer enters name and email
  4. System validates email format
  5. Once both fields are collected:
    • Client record is automatically created or linked
    • Conversation is linked to client
    • Customer can now send messages

Automatic Client Creation:

  • Client records created automatically when name and email are collected
  • Conversations are linked to client records for cross-channel tracking
  • System matches by email to prevent duplicate client records
  • Additional information (phone, company) extracted from messages and merged

Best Practices:

  • Ensure widget is configured to collect information before messaging
  • Review automatically created client records for accuracy
  • System prevents messaging until required information is collected

Session Management

Each chat conversation is linked to a session:

  • Unique session ID per visitor
  • Sessions persist across page loads
  • Multiple conversations per session (if widget reopened)

Real-Time Messaging

Chat conversations update quickly:

  • Messages are sent/received via the widget chat endpoint
  • The dashboard stays in sync via real-time tenant updates (WebSocket)

AI Responses

Chat messages receive AI responses:

  • Context-aware responses
  • Conversation history maintained
  • Assistant handoffs supported
  • Intent detection enabled
  • Customer name used for personalization (if available)

Managing Conversations

Search and Filter

Search:

  • Search by contact name, email, phone, or company
  • Results update as you type
  • Highlights matching text

Filters:

  • Channel: Filter by communication channel
  • Date Range: Filter by conversation date
  • Qualification: Filter by lead qualification status

Export

Export conversations to CSV:

  1. Apply desired filters
  2. Click Export button
  3. CSV downloads with all conversation data

Qualification

Mark conversations as qualified:

  • Qualified leads are ready for sales follow-up
  • Filter by qualification status
  • Track qualification metrics

Widget Integration

Customer Experience

From the customer's perspective:

  1. Widget appears on website
  2. Customer clicks widget button
  3. Chat window opens
  4. Customer types message
  5. AI responds immediately
  6. Conversation continues in real-time

Admin Experience

From the admin dashboard:

  1. Conversations appear in real-time
  2. View full conversation history
  3. Add notes and context
  4. Set AI conversation guides
  5. Export conversation data

Best Practices

  1. Monitor Conversations: Check conversations regularly
  2. Set AI Guides: Provide context for important conversations
  3. Respond Promptly: AI handles initial responses, but monitor for escalation
  4. Track Metrics: Monitor conversation volume and response times
  5. Export Data: Regularly export conversations for analysis

Troubleshooting

Conversations Not Appearing

  • Check WebSocket connection
  • Verify widget is sending messages
  • Review API logs for errors
  • Check conversation creation logic

Real-Time Updates Not Working

  • Verify WebSocket is connected
  • Check authentication token
  • Review browser console for errors
  • Test WebSocket connection directly

AI Responses Not Generating

  • Verify assistant is configured
  • Check default chat assistant is set
  • Review API logs for AI errors (model/provider misconfig, safety blocks, rate limits)

For more troubleshooting help, see Troubleshooting Guide.

autoch.at Documentation