Three itineraries that meet every constraint, ready to buy

What does the constraint math look like?

Checking that an itinerary satisfies "lands before 2 p.m.

local, no overnight layover, refundable, under cap" across every candidate is mechanical, and error-prone for a tired human at 11 p.m. A companion reads each result properly and does it for all of them without slowing down.

A fare quoted during research can change by checkout, and seat maps, baggage add-ons and fare-class fine print are worth your eyes. Three fully specified options with links (this flight, this fare class, this cancellation policy, this total) is ninety percent of the work already done.

Three people to the conference

Inputs: dates, the keynote deadline, two origin cities, hotel radius around the venue, refundability, a per-person cap. The companion searches the two origin routes across an aggregator plus the two airlines your team holds status with, checks each candidate against every constraint, then does the hotel pass for the surviving arrival times.

Output: three itinerary bundles with flights per person, hotel, total cost, cancellation terms, and the one trade-off per bundle stated plainly ("cheapest option lands 11:40 a.m., tight for a 2 p.m. keynote if delayed"). You pick a bundle, click through the provided links, and buy in ten minutes. Teams doing this monthly should read planning team travel; agencies live in a different world, see AI for travel agents.

Why is the tenth trip faster than the first?

Because you wrote the rules down once: you want the aisle, you avoid red-eyes unless dire, and you will pay extra to skip one specific airport. Every trip after that gets researched against those rules without you restating them.

Fares move between research and checkout, so the shortlist is worth acting on the same day. The companion browses at human pace in your own signed-in browser, including the airline accounts where your status and credits live, and hands you live links to click.

Frequently asked questions