TasteQA — Qloo-Powered Recommendation Auditor

Inspiration

Inspiration

Travel recommendations often prioritize popularity and proximity, but those factors do not always explain cultural fit. TasteQA was created to explore a more transparent approach: comparing location-based suggestions with recommendations supported by cultural-affinity data from Qloo.

What it does

TasteQA is a hospitality and travel recommendation auditing prototype. It compares a location-based baseline with Qloo-informed recommendations using cultural signals such as music, films, and brands.

The application helps users explore how cultural signals influence recommendations, where results differ, and where the available evidence is insufficient.

How we built it

The application uses JavaScript and a server-side API hosted on Vercel.

The backend connects to the Qloo API to resolve cultural entities and retrieve insights. The application compares those results against a location-based baseline and presents the differences through a public web interface.

The code is maintained in a public GitLab repository with an MIT license.

Challenges

A major challenge is distinguishing meaningful cultural evidence from incomplete or ambiguous signals. Cultural affinity does not guarantee individual preferences, and recommendations should not be presented as personalized facts without sufficient evidence.

Another challenge is keeping API credentials server-side while making the recommendation process understandable to users.

What we learned

TasteQA demonstrates why recommendation systems benefit from explainability, evidence boundaries, and comparisons against simple baselines.

It also highlights the importance of distinguishing population-level cultural signals from individual user preferences.

What's next

The next development milestone is to publish and independently verify an agent-callable Qloo tool with bounded execution, observable decisions, and safe failure handling.

The current public prototype demonstrates Qloo API orchestration. The additional agentic integration is not yet verified in the public deployment.

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for TasteQA — Qloo-Powered Recommendation Auditor

Built With

• css
• gitlab
• html
• javascript
• node.js
• qloo
• vercel