JoyRide.

JoyRide is an AI family travel planner built for busy parents juggling demanding careers and multiple kids across wide age ranges — collapsing the 8–15 hours of fragmented research across 4–8 sources (TripAdvisor, parenting blogs, Instagram saves, mom-tab chaos) into a single trustworthy itinerary that's age-aware, weather-informed, and geographically clustered.

The differentiator isn't any single feature — it's the combination of four: (1) Family-specific input (kid ages, optional gender, trip dates); (2) Weather-informed activity logic (season and forecast matching); (3) Geographic clustering (no zigzagging across the city); (4) Strict no-invented-venues / no-adult-venues rules with TripAdvisor "Travelers' Choice" badge integration. The promise is the deliverable: a clean, shareable itinerary instead of a Google Doc cobbled from 10 browser tabs.

Powered by a two-model Claude strategy: Claude Opus 4.6 for main itinerary generation (depth, planning quality, narrative); Claude Sonnet for ancillary activities and stores sections (faster, cheaper, sufficient quality). Started with Haiku for cost efficiency, then escalated to higher-tier models where output quality justified the cost. System prompt rules are strict: no invented venues, no restaurants/hotels invented, no adult venues, no specific times, no duplicate activities, age-appropriateness flags (toddler / kids / teens, stroller-friendly), and suggested duration per activity.

Targeting the family travel market in a perfect tailwind moment: 92% of parents plan to travel with kids in the next year (highest since the pandemic, per the 2025 Family Travel Association survey), 100M+ Americans expected to take family vacations, and AI in tourism growing at 28.7% CAGR through 2030.