Why We Built Longtable.
May 23, 2026
Three families. A group chat that had been going for six years. The kind of thread where someone drops a link to an Airbnb at 11 p.m. on a Tuesday, three people heart-react, and then nothing happens for two weeks. Eventually someone texts “so are we doing this?” and no one answers because no one knows what “this” is yet.
There was a shared Google Doc that one partner had started. It had four destination ideas, two of which were jokes. The budget column was empty. The dates column said “October?” — a question, not an answer. Someone had opened a Doodle poll, but only two of the six adults filled it out. Six weeks later, the trip didn’t happen. Not because anyone said no. Because no one could get past the starting point.
The trip died in the gap between “we should do something” and “here are three real options.”
The fragmentation problem
The idea lives in iMessage. The research lives in open browser tabs. The dates live in a Doodle poll half the group forgot. The budget lives in someone’s head. The itinerary lives in a Google Doc nobody knows is the real one. By the time you’ve checked four apps and three threads, the planning energy is gone.
No single tool owns the workflow from idea to itinerary. Each one handles a slice — polls, docs, booking, payments — but none of them connects the slices. The planning energy dissipates across all of them, and the trip dies not from conflict but from friction. The intent was there. The starting point wasn’t.
The household, not the individual
Every planning tool treats each person as a user. But a trip with three families isn’t shaped that way. One partner plans; the other wants to be told when to show up. The kids have constraints the parents forget to mention until the weekend is almost locked. The budget is per household, not per person. The calendar bottleneck is per household — two adults sharing one schedule, not two independent people.
The product should reason over the intersection of household constraints, not individual preferences. When each household submits once — calendar windows, budget ceiling, kid logistics, vibe preferences — the feasible overlap reveals itself. The partner who never opens the doc becomes a first-class user, not an afterthought. Their silence is data, not neglect.
A co-planner that does the research
Type a paragraph: “fall weekend, three families, four hours from LA, around fifteen hundred a family.” Three real options come back in five minutes. Not generic AI filler. Specific destinations. Specific lodging with pricing that fits the budget. Specific activities anchored to the place. Honest reasoning about the trade-offs — if one option exceeds budget by $200 but has the best lodging, Otto says so.
Otto is a two-phase research engine built on Claude, Anthropic’s frontier AI models. Phase one maps the possibility space — destinations that fit the constraints, dates that clear every household’s calendar. Phase two synthesizes each candidate into a full trip card: lodging within budget, anchor activities, restaurants, estimated cost per household. The output is not a list of links. It’s three ready-to-decide options.
The planning artifact exists before the group decides whether to plan. That’s the unlock.
One workflow, start to finish
Kickoff. Research. Constraint solving. Date lock. Itinerary. Memory card. Plan another. One workflow covering the full lifecycle of a multi-household trip, in one product. The source of truth lives in one place, not scattered across four apps and three threads. Decisions made are recorded. Decisions pending are surfaced. The itinerary lands in everyone’s calendar a week before. The memory card arrives a few days after.
What’s next
Calendar sync, so the date lock writes itself into everyone’s Google Calendar. Booking integration, so the lodging goes from “selected” to “reserved” in one click. The trips-per-year flywheel — after every trip, a memory card surfaces six months later with one prompt: plan another? If the friendships matter, the answer is usually yes, and the next trip starts with the same households already in place.
A tool that earns its place in your calendar twice a year, for fifteen years. That’s the ambition.
The trips that almost didn’t happen — those are the ones that matter most. Longtable exists to make sure they happen.