All use cases

Prevent / Care

Flight-delay proactive rebooking

A traveler's flight just slipped and their connection is at risk — the agent reaches them with rebooking options before they ever check the departures board.

Get aheadSystem triggerRisk triggerWhatsAppPushSMSTravel

What it does

When a flight is delayed, most travelers find out at the gate or deep in a support queue, already furious and already late. For a traveler with a tight connection, every minute of silence makes the recovery worse and the eventual support contact angrier. The agent gets there first: the moment the delay fires from ops, it weighs what it knows about this specific traveler — a tight onward connection, a previous delay that went badly, a preference for WhatsApp — and sends a quality-gated message proposing concrete later options that protect the journey. The traveler arrives informed instead of furious.

How it works

  1. 1
    Trigger. a flight_delayed event arrives from the disruption feed, carrying the new estimate and the affected itinerary.
  2. 2
    Decision. the agent maps the delay against the traveler's itinerary and memory — does this break a connection, how did past disruptions land with this person, which channel do they actually answer on — and selects viable later options from the reservation system. The judge clears the message before anything sends.
  3. 3
    Action. an outbound on WhatsApp, push, or SMS that names the delay plainly and offers later options that protect the connection, with a one-tap path to rebook.
  4. 4
    Follow-through. when the traveler confirms, the rebooking executes through the operator's own reservation flow — the agent proposes, the client's systems execute. If the delay is rolled back or the traveler rebooks on their own first, the now-stale outreach is cancelled. The exchange writes back to the traveler's memory for the next disruption.
Example
AgentYour 14:20 is delayed. Two later options still protect your Lisbon connection — want me to rebook?
CustomerYes, the earlier of the two please.
AgentDone — your rebooking request is in and your new confirmation is on its way. Your connection is protected, and I'll message you here if anything else changes on this trip.

Configuration

How the agent is wired for this use case.

Triggera flight_delayed event from the disruption feed, carrying the new estimate and the affected itinerary
Tools & actions
  • Disruption feed · ingest the delay and schedule-change event, map it against the affected itinerary
  • Reservation system · read viable later options that protect the connection
  • Reservation system · execute the confirmed rebooking through the operator's own flow
  • Messaging channel · send the quality-gated outreach with a one-tap path to rebook
  • Memory store · write the exchange back for the next disruption
Autonomydetection, option selection, and the judge-gated send run unattended; the rebooking only executes on traveler confirmation — the agent proposes, the client's reservation flow executes, nothing is ticketed without a yes
Channelswhatsapp · push · sms
Escalationstale outreach is cancelled if the delay rolls back or the traveler rebooks first; anything outside the proposed options stays with the operator's reservation flow

What you need

The inputs this use case runs on. Your channels stay yours; the agent supplies the judgment.

Signals

flight_delayed and schedule-change events from the ops/disruption feed; itinerary and connection data per booking

Data

traveler itinerary with connection windows, past disruption history, channel preference, consent state

Guardrails

judge gating on every unprompted message; the agent proposes and the client's reservation flow executes, nothing is ticketed without confirmation; stale-outreach cancellation when the situation changes; frequency caps per disruption

Metrics it moves

  • contact-ratedown, because the traveler who already has options in hand never joins the angry support queue
  • ticket-deflectionup, with prevented disruption tickets attributable to the outbound that landed first
  • time-to-resolutiondown, since the first touch happens ahead of the complaint instead of after it
  • csatup, because being told early with a way out is the single biggest sentiment swing in a disruption

Related use cases

Prevent / Care

Delivery-trouble proactive notice

the same get-ahead pattern on parcels instead of flights

Prevent / Care

Known-incident broadcast with goodwill

one incident, many affected customers, same lead-with-the-news instinct

Prevent / Care

Preventive CX — reach out before the user arrives angry

the general form of reaching out before the anger forms

See it on your own customer journey

Bring one drop-off, one churn cliff, or one silent segment. We will show you what a proactive agent with memory and judgment does with it.

Book a demo