Step one · the family

Everything you mean to show them, on a calendar that knows when they leave.

An advisor runs this with the whole family in the room. Two minutes here; in practice it is an hour-long conversation and a much longer list.

Who's going

Ages drive everything. Most kids remember little before about seven, and most stop wanting to come somewhere around sixteen. That is the window the plan is solving inside — not the horizon you picked.

The shape of it

Travel budget per year

45,000 / year

Plan through

Three trips a year is the default rhythm: spring break, a longer summer, and something over winter break.

The list — everything you mean to do

Pre-loaded with a plausible family's list. Click to add or drop. 14 selected.

Nothing is sent anywhere. This runs in your browser.

The window — who is still home, year by year

Step two · the plan

The calendar

Every trip carries the reason it sits where it sits. That line is the product.

Step three · the part nobody else can do

Something changed.

A trip-by-trip advisor can cancel or postpone. An advisor holding ten dated trips can rearrange, because there is always another trip whose season fits. You stop losing the year.

What moved

What this is, and what it isn't

The data is hand-built: Forty destinations with season windows, rough costs for a family of four, and a sensible minimum age — assembled by hand for this demo, not pulled from a source. The seasons are real (peak blossom really is a two-week window in late March; the Serengeti migration really does run July to October). The costs are estimates.
The solver is deliberately simple: It walks the calendar in order and, for each open slot, takes the highest-scoring trip that fits the season, the budget, the days available, and how old the kids are that year. Urgency wins ties — a trip that expires when the oldest turns eighteen gets placed before one that will still be there in a decade. A real version would weigh a great deal more.
Nobody has been asked for money: This shows the artifact an advisor would sell. It does not prove she can sell it, and that is the venture's biggest open question.
Almanac is a working name: The software trademark is dead and nothing blocks it, but almanac.io is a live company in another category. A proper naming pass comes later.