How Otto Plans a Trip.

Someone types a paragraph into the kickoff box: “fall weekend, three families, four hours from LA, around $1,500 per family.” That’s the entire brief. No form fields, no dropdown menus, no “select your preferred destination from a list of 200.” One paragraph. Otto takes it from there.

Phase 1: The Planner

Otto’s first phase maps the possibility space. From that paragraph, it extracts the constraints — group size, budget ceiling, drive radius, time of year — and evaluates a dozen candidates against them. Sedona: five and a half hours, but October weather is perfect and the lodging fits the budget. Big Bear: two hours, but cabins in October book early and the good ones are already taken. Santa Barbara: ninety minutes, coastal shoulder season, hotel rates drop after Labor Day. Joshua Tree: two and a half hours, but thin on kid-friendly infrastructure.

The planner doesn’t just list destinations. It reasons about fit. Driving hours from LA. Seasonal weather patterns. Whether lodging at the stated budget actually exists. Whether the vibe matches a group with kids versus a group without. Four candidates survive. The rest get cut with a sentence explaining why.

Phase 2: The Synthesizer

Three finalists get enriched into full trip cards. For Sedona, that means specific lodging — L’Auberge de Sedona at $495 a night, Enchantment Resort at $425, Amara Resort at $329 — with notes on hot tubs, proximity, and whether the pool is kid-friendly. Named activities with duration estimates: Cathedral Rock trail (two hours), Devil’s Bridge hike (two and a half), Slide Rock State Park (two hours with the kids). Restaurant recommendations: Elote Cafe, Mariposa, The Hudson. Cost per household calculated: roughly $1,650.

Every detail is specific. Every price is researched. Every recommendation comes with reasoning — if one lodging option exceeds budget but has the best location, Otto says so. If an activity is a 45-minute drive from the hotel, that’s noted. The synthesizer doesn’t generate filler. It builds a trip you could actually take.

The output: three cards

What comes back is not a list of suggestions. It’s three ready-to-decide trip options. Sedona, Joshua Tree, Santa Barbara. Each card shows the destination, the vibe in one sentence, lodging with real pricing, three anchor activities, and the estimated cost per household. Side by side, comparable. The group picks one — or picks none and Otto re-runs with adjusted constraints. Either way, the planning artifact exists before the group decides whether to plan. That’s the unlock.

The gap between “we should do something” and “here are three real options” — that’s what Otto closes.

What makes this different from ChatGPT

You can ask ChatGPT to plan a trip. It will give you a cheerful list of destinations with generic descriptions pulled from travel blogs. No pricing. No lodging availability. No awareness of your budget or group size. No structured output you can compare side by side. No household-level reasoning.

Otto doesn’t generate filler. Every place is real. Every price is researched. Every recommendation comes with reasoning. The output is structured — three cards, each one a trip you could actually take. Not inspiration. Decisions.