Every way a meeting can happen.
Scheduling looks simple from the outside — pick a time, send an invite. In practice, a single request can vary across dozens of independent dimensions at once: who’s involved, who represents them, what time zones they’re in, what kind of meeting it is, what can change after it’s booked.
A hybrid all-day offsite for 22 people across 3 offices and 4 time zones — catering and AV booked per session, one VP's EA holding override authority to bump a lower-priority hold.
An 11-person deal review across 2 companies and 4 time zones, 3 of whom have EAs, rescheduled twice after a cancelled flight, resolved through parallel human and AI negotiation.
A client dinner for 6, blended business and social, reservation-bound — one attendee's flight lands 40 minutes before start and cascades a reshuffle of the pre-dinner call.
A recurring weekly sync with a rotating attendee list, where one recurring guest now represents themselves through their own AI assistant instead of attending directly.
None of these axes apply one at a time — a single meeting sits at the intersection of all of them at once. Multiply them out and the space of possible scheduling scenarios runs into the billions. No two requests are really the same, and our engine is built to handle the full matrix, not just the common cases.
This isn't a category, it's a multiplier — every additional required attendee shrinks the feasible window and adds one more set of constraints that must all be satisfied at once. Complexity compounds, it doesn't add.
18 dimensions, one scheduling engine.