Inspiration

Every group trip dies the same way: someone drops an idea in the chat, everyone reacts with enthusiasm, and then it sits there for months while people "check their calendars." I wanted to build the opposite of that — a commitment device, not another planning tool. Instead of giving groups more time to deliberate, Send It gives them less: a real, price-outlier destination and a three-minute window to say yes or watch it disappear.

What it does

Send It is a FOMO-driven group travel app built around a single mechanic: The Drop.

  • Users create or join a Group with an invite code.
  • Once a week, at a random time, a Drop is sent out, and the app scans real accommodation pricing across destinations to surface a genuine price outlier — a city that's actually cheaper than its usual median right now.
  • The app generates a destination image, a high-energy AI-written pitch, and a sustainability score (0–100) with a weighted breakdown (travel footprint, grid cleanliness, transit access, local character).
  • Everyone in the group sees the same live card at once, with a 3-minute countdown ring that turns urgent red under 30 seconds.
  • The group votes Send It 🔴 or Nah.
  • Hit the vote threshold and the trip locks — a deposit and transaction record are created. Let the timer run out and it's a miss.
  • Every drop, lock, and miss is saved to a History view, and a shareable landing page (/drop/:id) lets a specific drop be sent directly to a friend.

How its built

  • Frontend: React + Tailwind, styled with a warm, editorial, anti-corporate aesthetic
  • Backend/platform: Base44 for entities, backend functions, and integrations.
  • Real data: Stay22 for live accommodation listings, pricing, and booking links, so the price-outlier detection is based on real median pricing rather than mock data.
  • AI layer: Gemini-grounded LLM calls generate the destination pitch copy and the sustainability score/breakdown.

Challenges

  • Getting a genuinely live-feeling countdown and vote mechanic working reliably across simulated and real participants at once.
  • Making the sustainability score feel trustworthy rather than decorative — grounding it in real signals and flagging it as an estimate when data was thin, instead of just guessing a number.
  • Balancing "FOMO and urgency" in the copy without the app feeling gimmicky or untrustworthy — a lot of iteration went into the tone of the pitch copy and the vote screens.

Accomplishments

  • Pairing a fun, urgency-driven mechanic with real underlying data (real pricing, a real sustainability score) instead of building pure vaporware FOMO.
  • A distinct, considered design voice — most travel/booking demos look like generic SaaS; Send It has actual personality in its copy and visual identity.

What I learned

  • How to design a real-time, multi-user interaction (the vote-and-lock mechanic) on a tight timeline.
  • How to combine a grounded LLM call with real third-party pricing data to produce something that reads as trustworthy.

What's next for Send It

  • Real payments/deposits instead of simulated ones, with an actual booking flow through Stay22.
  • Expanding price-outlier detection beyond Europe.
  • Real push/SMS notifications to group members when a drop fires (the SMS integration is already wired, just stubbed).
  • A "streaks" or group history leaderboard to reward crews that actually lock trips.

Built With

Share this project:

Updates

Submission history