Appmuseme

AI-Powered Travel Itineraries Based on Your Vibe

Comment

Inspiration

Most travel apps tell you where to go based on popularity, reviews, or influencer trends — but none ask what actually moves you. We wanted to flip that: what if you could plan a trip based on your cultural tastes instead of Instagram checklists?

Appmuseme is inspired by the idea that everyone travels differently. A jazz-loving vintage shopper shouldn’t have the same Tokyo experience as a techno-obsessed foodie — and with cultural AI, they don’t have to.

What it does

Appmuseme is a taste-powered travel assistant. It builds smart, culturally aligned travel itineraries based on what you love — not what's trending.

You just tell it where you're going and what you're into — music, food, mood, fashion, even favorite books — and Appmuseme returns:

A multi-day itinerary with places and experiences matched to your tastes

Local venues, cafes, bookstores, neighborhoods, and events you’d actually love

Optional extras like playlists, local authors, or style guides

It’s powered by a custom multi-agent system and Qloo’s Taste AI, which connects your preferences across domains like music, fashion, dining, and travel.

How I built it

Frontend: Next.js + Tailwind CSS for a clean, mobile-friendly UI

LLM Layer: Google Gemini (via Vercel AI SDK) parses freeform taste input from the user

Queue System: Routes user data from the frontend to the backend asynchronously with Quirrel.dev

AppmusemeAgent is the root agent

API: Qloo’s Taste AI to power cross-domain taste intelligence

Challenges we ran into

Designing prompts that could translate vague, creative inputs into structured data without losing nuance

Mapping Qloo’s cross-domain data into coherent daily itineraries — it’s powerful, but requires thoughtful orchestration

Integrating Gemini and backend smoothly via a queue system

Synthesizing taste signals (like “dream pop” or “brutalist cafes”) into local recommendations with cultural depth, not noise

Accomplishments that we're proud of

Built a full-stack, cross-system AI experience from scratch in a short time window

Designed a taste-to-itinerary pipeline that actually feels personal

Created a scalable multi-agent architecture ready to support more domains

Turned an abstract idea — "travel by taste" — into a real, working prototype with strong product potential

What we learned

  • How to orchestrate multi-agent systems with real-world APIs and LLMs
  • The importance of prompt design when working with cultural inputs -That travel is deeply personal — and AI can do more than generic recommenders if trained on the right taste graphs
  • How to turn deep data into user-friendly magic with the right interface

What's next for Appmuseme

  • User accounts to save and reuse taste profiles for future trips
  • Collaborative trip planning — merge tastes from multiple friends into one plan
  • Reverse discovery: Suggest new cities that match your vibe
  • iOS/Android version for on-the-go experience syncing
  • Partnerships with travel brands or cultural curators
  • City-wide taste maps that visualize local vibes for nomads, creatives, and explorers

Built With

Share this project:

Updates