Start
  ↓
Greeting
  ↓
Identify Intent → Appointment? → Collect Details → Book → Confirm
                → Question? → Search KB → Answer
                → Transfer? → Route Call
                → Other → Fallback

// Intent-based
if (intent === "book_appointment") {
  goto("collect_appointment_details");
}

// Keyword-based
if (userInput.includes("emergency")) {
  goto("escalate_to_human");
}

// Data-based
if (customer.hasAppointment) {
  goto("existing_appointment_flow");
}

{
  "id": "appointment_booking",
  "name": "Appointment Booking Flow",
  "nodes": [
    {
      "id": "start",
      "type": "start",
      "next": "greeting"
    },
    {
      "id": "greeting",
      "type": "message",
      "content": "Hi, I can help you schedule an appointment. What day works best for you?",
      "next": "collect_date"
    },
    {
      "id": "collect_date",
      "type": "listen",
      "variable": "appointment_date",
      "next": "collect_time"
    },
    {
      "id": "collect_time",
      "type": "message",
      "content": "Great, and what time would you prefer?",
      "next": "listen_time"
    }
  ]
}

Listen → Classify Intent
              ↓
    ┌─────────┼─────────┐
    ↓         ↓         ↓
Schedule  Question  Transfer

Ask Name → Store → Ask Phone → Store → Ask Email → Store → Confirm All

Detect Escalation Trigger
        ↓
Check Agent Availability
        ↓
    Available? ─→ Warm Transfer
        ↓
    Unavailable → Collect Callback Info → Promise Follow-up

User Question → Search KB → Found? → Deliver Answer
                              ↓
                           Not Found → Apologize → Offer Transfer

// Set variables
context.customerName = "John";
context.appointmentDate = "2024-01-15";

// Use in messages
"Great, {{customerName}}, I've booked you for {{appointmentDate}}."

Development → Test → Staging → Validate → Production