TastePassport

Same city. Different trip. Taste-grounded 1-day travel itineraries via the Qloo taste graph.

Comment

Elevator pitch

Same city. Different trip. TastePassport turns your taste — the artists, films, foods and brands you love — into a 1-day travel itinerary grounded in Qloo's taste graph, and proves it side by side against the generic tourist trip.

Inspiration

Two friends land in the same city with the same guidebook and get the same boring day. One loves indie film and ramen; the other lives for street food and hip-hop. The guidebook serves neither. Travel recommendations are generic because they know nothing about who you are. Qloo's taste graph knows what people love — we built the trip planner that finally uses it.

What it does

Pick 3–5 favorites (music artist, film, cuisine, brand) or tap a preset like "Indie film lover" or "Street food hunter", pick a city, and TastePassport builds a 1-day itinerary: Morning / Afternoon / Evening, each stop with a one-line "Why this fits" taste reasoning and its Qloo entity type. A Taste DNA strip shows the inferred tags behind the plan. Then the killer demo: the generic tourist trip vs your taste trip, side by side, for the same city — or two taste profiles compared head to head. The same city gives two completely different days, because taste travels.

How we built it

Single-file HTML + CSS + vanilla JavaScript (zero install, GitHub Pages). Demo mode runs on a curated dataset of 40 recommendations scored by taste-tag overlap, so judges can try it with no key. Live mode calls Qloo's Insights API (GET /v2/insights, filter.type place / movie / artist / brand) with a user-supplied hackathon key stored only in browser localStorage. Any API error falls back silently to demo data with an honest "demo data" badge — the demo never breaks.

Challenges we ran into

Qloo's Insights API silently ignores parameters that are not valid for the chosen filter.type, returning an empty list with no error. We learned to validate every parameter against the entity-type guide and to treat the curated demo dataset as the reliability backbone, not a placeholder.

Accomplishments that we're proud of

The side-by-side comparison makes the value of cultural intelligence visible in one scroll: same city, different taste, different trip. And the app never fails the demo — live Qloo data when the key is present, honest demo data otherwise.

What we learned

Cultural intelligence is the difference between "technically functional" and "actually good." A recommendation engine that knows what you love does not just rank better — it changes what a day in a new city feels like.

What's next for TastePassport

More cities and multi-day trips, group mode (merge a whole travel party's tastes into one plan), and saved taste passports you can reuse across trips.

Built with

HTML, CSS, JavaScript, Qloo Taste AI API

Built With

Share this project:

Updates

Submission history