All work

Case study · 04 / 04

Roamly

A full-stack travel discovery platform combining real traveler ratings with an AI chat assistant.

Roamly screenshot

Problem

Planning a trip usually means sifting through generic listings with no trustworthy signal. Roamly solves the decision fatigue by grounding recommendations in real traveler data.

Approach

Destinations are backed by real trip ratings, and an AI assistant powered by Google Gemini gives personalized recommendations. Google OAuth via BetterAuth, a live rating system with real time data fetching, and fully independent frontend and backend deployments complete the stack.

Key Decisions

  • Resolved cross domain cookie authentication in production with a Next.js rewrite proxy, so every API call appears same origin to the browser.
  • Integrated the Gemini API to power a conversational assistant that gives genuinely context aware travel suggestions.
  • Replaced an initial recommendation engine concept with a trip rating system for a more trustworthy, user driven signal, based on clarified requirements.

Outcome

A travel discovery platform where user submitted ratings and an AI assistant both feed into the same recommendation experience.

What's Next

  • Collaborative trip planning where multiple users build an itinerary together.
  • Personalized destination feeds based on past ratings and chat history.
  • Expanding the assistant to generate full day by day itineraries.

Next case study

Brawnix