Integrations

AI Auto-Responses Guide

Use AI to draft responses, classify tickets, and handle common inquiries automatically.

Auto-Draft Replies

AI generates response drafts for agent review and approval.

Smart Classification

Automatically categorize and prioritize incoming tickets.

Auto-Tagging

Apply tags based on ticket content and intent.

24/7 First Response

Acknowledge tickets instantly, even outside business hours.

AI Use Cases

Ways to leverage AI in your support workflow.

FAQ Auto-Response

Automatically answer common questions using your knowledge base.

Example: Password resets, shipping times, return policies

Sentiment Detection

Flag frustrated customers for priority handling.

Example: Urgent escalation for negative sentiment tickets

Intent Classification

Route tickets based on detected intent.

Example: Billing questions → Finance team, Bugs → Engineering

Response Suggestions

Draft responses for agents to review and personalize.

Example: Agent approves or edits AI draft before sending

Build Your Own AI Integration

Use webhooks and our API to add AI capabilities.

1. Listen for New Tickets

// Set up webhook for new tickets
POST /v1/workspaces/ws_xxx/webhooks
{
  "url": "https://your-ai-service.com/process",
  "events": ["ticket.created"]
}

2. Process with AI

// Your AI service receives the ticket
{
  "event": "ticket.created",
  "ticket": {
    "id": "tkt_xxx",
    "subject": "Can't reset my password",
    "body": "I've tried clicking the reset link..."
  }
}

// Analyze with OpenAI, Claude, etc.
const analysis = await openai.chat.completions.create({
  model: "gpt-4",
  messages: [
    { role: "system", content: "Classify this support ticket..." },
    { role: "user", content: ticket.body }
  ]
});

3. Update Ticket with AI Results

// Add classification and draft response
PATCH /v1/workspaces/ws_xxx/tickets/tkt_xxx
{
  "priority": "normal",
  "tags": ["password-reset", "faq"],
  "custom_fields": {
    "ai_category": "account_access",
    "ai_sentiment": "neutral",
    "ai_draft_response": "Hi! Here's how to reset your password..."
  }
}

// Optionally auto-respond for FAQs
POST /v1/workspaces/ws_xxx/tickets/tkt_xxx/comments
{
  "body": "Hi! Here's how to reset your password...",
  "author_type": "agent",
  "auto_response": true
}

Best Practices

1

Start with drafts, not auto-send

Let agents review AI responses before sending to build trust and catch errors.

2

Use AI for classification first

Routing and tagging are lower-risk ways to get value from AI.

3

Train on your knowledge base

Better AI responses come from context about your product and policies.

4

Monitor and iterate

Track AI accuracy and continuously improve your prompts and rules.

Add AI to Your Support Stack

Respond faster and scale your support with AI.