dnata
Control Tower
Powered byAgentic Fluxus

Agent Settings

Human-in-the-loop configuration & processing rules

Category Review Rules

Configure which complaint categories require human review before sending

Baggage — Standard Claims

Claims under €500 — AI can auto-resolve

AI Auto

Baggage — High Value Claims

Claims over €500 — always requires human review

Human Review

Cargo Damage

Insurance implications — always human review

Human Review

Catering Quality

Standard catering complaints — AI can auto-resolve

AI Auto

Safety Incidents

Legal and regulatory — always human review

Human Review

Ground Equipment

Investigation required — always human review

Human Review

Positive Feedback

Auto-acknowledge with brand voice

AI Auto

Booking Amendments

Standard corrections — AI can auto-resolve

AI Auto

Brand Voice Confidence

Responses below this confidence score are automatically flagged for human review

85%

Currently: responses below 85% brand match are held for review

General Settings

Require Human Approval for All

Override: every response must be manually approved

Auto-Escalate Repeat Senders

If customer replies 3+ times, escalate to human

Queue Outside Working Hours

Responses queued for morning review (06:00–22:00 CET)

Email Notifications

Send email alerts for escalations and urgent items