Booking Revenue Anomaly Detection

osop_version: "1.0"
id: "booking-anomaly"
name: "Booking Revenue Anomaly Detection"
description: "AI detects unusual booking patterns and alerts revenue managers for pricing adjustments."

nodes:
  - id: "booking_stream"
    type: "event"
    name: "Booking Event Stream"
    description: "Real-time feed of reservations, cancellations, and modifications."

  - id: "revenue_analysis"
    type: "agent"
    subtype: "llm"
    name: "AI Revenue Pattern Analysis"
    description: "Analyze booking velocity, ADR, and occupancy against forecasts."

  - id: "anomaly_detect"
    type: "system"
    name: "Anomaly Classification"
    description: "Flag demand surges, unusual cancellation spikes, or rate parity issues."

  - id: "manager_alert"
    type: "api"
    name: "Revenue Manager Alert"
    description: "Dashboard alert with anomaly details and recommended action."

  - id: "pricing_decision"
    type: "human"
    subtype: "review"
    name: "Adjust Pricing Strategy"
    description: "Revenue manager reviews AI recommendation and adjusts rates or inventory."
    security:
      approval_gate: true

edges:
  - from: "booking_stream"
    to: "revenue_analysis"
    mode: "sequential"
  - from: "revenue_analysis"
    to: "anomaly_detect"
    mode: "sequential"
  - from: "anomaly_detect"
    to: "manager_alert"
    mode: "conditional"
    when: "anomaly.detected == true"
  - from: "manager_alert"
    to: "pricing_decision"
    mode: "sequential"
  - from: "revenue_analysis"
    to: "manager_alert"
    mode: "timeout"
    timeout_sec: 120
    label: "Escalate if analysis exceeds 2min"