Inspiration

One-person recommendation engines are not designed for shared decisions. CommonGround tackles the more interesting problem: finding culturally grounded experiences that two people with different interests may both enjoy.

What it does

Two people enter a few favorite movies, artists, shows, books, or brands and select what they want to discover together. The agent resolves the favorites to Qloo entities, retrieves individual and combined recommendations, then assigns a fairness-oriented balance index and provides explanations.

How we built it

  • Responsive browser experience with HTML, CSS, and JavaScript.
  • Serverless Node.js orchestration on Vercel.
  • Real Qloo entity lookup via GET /search and cultural candidates via GET /v2/insights.
  • Three independent Qloo Insights requests (person A, person B, and combined interests), followed by fairness reranking.
  • Server-only QLOO_API_KEY, stored as a protected Vercel production secret.

Why Qloo is essential

Recommendations need Qloo's cross-category cultural intelligence, not a generic language model guessing what fits. In live mode, the candidates come from Qloo; the balance index is our own transparent heuristic using individual and combined result ranks.

Challenges and what we learned

Cross-person discovery requires genuine cultural entity resolution, cross-domain signals, and a ranking objective that does not simply reward the most popular answer. We also made a sharp distinction between verified results and fictional preview examples.

Implementation status — November 1, 2026

The public app and repository are deployed. The hackathon API key has been configured as a protected production environment variable and a fresh Vercel production deployment reports READY. A full end-to-end Qloo response has not yet been independently verified, so we do not claim demonstrated recommendation accuracy. Preview mode remains clearly labeled and synthetic.

Links

Live app: https://commonground-qloo-2026.vercel.app Source (Apache-2.0 licensed parent repository): https://github.com/samertall/tools/tree/main/qloo-commonground

Author: Samer Tallauze.

Built With

Share this project:

Updates

Submission history