Crew Pairing Optimization demo
Builds the month's legal, lowest-cost crew trips with trade-offs shown, for pairing analysts 1:32
What the demo shows
Builds the month's legal, lowest-cost crew trips with trade-offs shown, for pairing analysts
The month's candidate solutions sit on one frontier — cost against crew quality of life, with the share of pairings that hold on replay ringed on every point — beside the pairing agents. A schedule reader diffs the month and asks network planning about doubtful flights; a rules keeper turns Part 117, the contract and new letters of agreement into rules with their source passages; a pairing builder runs dozens of solutions and re-runs on a plain-words request; an independent legality and fatigue checker re-tests every pairing; an operation replayer flies each solution against 730 days of real delays; a trade-off explainer writes one sentence per point with citations. The analyst chooses, the scheduling committee reviews, the director approves with a reason code and e-signature, and a publish agent re-verifies before anything reaches bidding.
Transcript
- Crew Pairing Optimization builds an airline's monthly crew trips for pairing analysts — legal and low-cost.
- Pay is a big U.S.
- carrier's largest cost — $16 billion a year, much of it crew.
- Overnight, agents built 46 solutions — cost against crew quality of life.
- Solution 3 costs $41.2 million, 4 adds 12 Tokyo layovers.
- Megan asks for fewer mid-trip red eyes, under 41.4.
- The builder reruns. A checker retests every pairing.
- Solution 8. 46 fewer red eyes, 100,000 more.
- This trip had a red eye, then 10 hours rest.
- Now it ends at home.
- Megan submits it, and the union committee reviews.
- The director signs with a reason.
- An agent re-verifies and publishes.
- Every number cited, every decision signed, every step logged.
- $800,000 below last month's method, every pairing legal.
- Built on Botminds: agents, data, screens, and governance on one platform.
Transcribed from the film’s narration. Captions are on the player.