Accenture + OpenAIQantas booking companion · concept demoReference preview · live connections pending
AIRLINE INTELLIGENCE

A clearer way
to choose your flight.

A Qantas flight-buying experience: inspiration helps choose where to fly, then grounded fare guidance helps the traveller compare and take the next step.

Text conversationRealtime voiceTraveller-supplied preferences
BOOKING COMPANION · CHAT + VOICE

OpenAI orchestration

Flight intent · fare trade-offs · a clear next step

01 / AIRLINE KNOWLEDGE

Qantas information

Current fare conditions, applicable loyalty rules and flight-related benefits, supported by official source links. No invented service or points claims.

02 / FLIGHT OFFERS

Qantas-issued fare search

The Duffel feed compares traveller-confirmed flight legs in one request. Returned offers identify the itinerary and carriers; a Qantas-issued offer is not necessarily Qantas-operated.

03 / TRAVELLER CONTEXT

Preferences for this visit

Guest-entered trip details or a clearly labelled illustrative member example. No customer-record lookup or authenticated loyalty balance.

THIS DEPLOYMENT

Connection readiness

Configuration presence is shown below. A configured connection still needs an end-to-end check before a live presentation.

OpenAI · Studio, chat + Realtime voiceConnection pending
Duffel Qantas fare searchConnection pending
FlightAware adapterConnection pending
THE QANTAS PATH

Build towards the flight purchase.

  • Confirm airports, dates, travellers and cabin before checking fares.
  • Compare matching offers, operating carriers and actual fare conditions.
  • Verify loyalty eligibility separately; cash fares are not reward inventory.
  • Confirm the final offer and complete the purchase on Qantas.

Route ideas are not available seats.

Journey Studio uses destination research to help choose arrival and departure gateways, the order of a trip and time on the ground. Its itinerary is supporting context for flight choices, not a hotel, activity or holiday-package booking service. A route idea does not prove that Qantas serves it or that a fare is available.

Trip Preview links to a separate prepared family summary. A fixed scenario link opens Alex in Studio: selected, unpurchased flights, a Sydney itinerary and 120,000 illustrative starting points. The link passes only an allowlisted scenario identifier. It does not transfer a private ChatGPT conversation, sign into a loyalty account or synchronize edits between the two sites.

The points-and-cash preview uses a bounded local calculation with a disclosed, made-up pricing schedule. It shows the whole-party USD cash amount and remaining example points. It is not a live Qantas quote, redemption rate, reward-seat search or payment action. Voice and text use the current Studio context; live planning and fare searches remain separate, explicit requests.

Structural checks validate night totals, complete day sequences and overnight-base allocations. They do not verify airline schedules, through-ticket protection or availability. Flight details need the traveller’s confirmation before a fare check, and any returned fare remains a snapshot to verify before purchase.

Studio drafts, captions, the points preview and up to three kept versions last for this page visit. Reset Alex demo restores the prepared baseline and clears transient work; leaving or reloading does not preserve edits. Its planning history is separate from the booking conversation; the traveller reviews the flight details and any handoff before proceeding. Nothing is purchased automatically.

Member examples are illustrative. This experience cannot authenticate loyalty balances, redeem points, write customer cases, take payment or issue tickets. The original Connected Journey deployment remains separate and unchanged.