layover
- AWS Lambda
i built layover because planning trips with my friends was getting harder than it needed to be.
Layover helps a group of friends decide where to go, in what order, for how many nights, and what it will all cost. Everyone says what they want and what they can spend, and it works out which trips are actually possible.
My friends and I are always talking about trips we want to take, but once we actually start planning, it gets messy. Everyone has different budgets, dates, and places they want to go. We end up trying to keep track of everything in a group chat, where details get buried and it's hard to know what's been decided. I wanted a better way to organize it all.
All of it. I decided what the product should do, designed the data model and the planner, built the app, and set up the cloud pieces around it.
The planner is deterministic. It computes routes, feasibility and budgets itself, and the AI only explains the results and helps people navigate the trade-offs. That split means it can't invent a price or quietly override a constraint.
Private budgets stay private because the database enforces who can see what, not just the interface. A small cloud worker fetches flight pricing on a schedule so the planner works from real numbers.
The scoring and a few of the screens have been redesigned more than once, and some still feel unsettled. Next I want to watch real groups use it and let that decide what to simplify.