Overview
OmniPulse × Qloo is a taste-aware experience agent. It turns a user's explicitly selected cultural references into a coherent local experience grounded in Qloo's cultural intelligence graph.
How it works
- A user enters up to three cultural references such as a film, artist, book, brand or game.
- The backend calls Qloo's entity search endpoint and shows candidate matches for the user to confirm.
- The user confirms the exact Qloo entity IDs. Those IDs become the explicit taste signals for this session.
- The agent calls Qloo Insights to find affinity-ranked places in the selected city, plus related films and artists.
- OmniPulse composes the results with an occasion-specific ambient-scene preview and displays a trace of the Qloo calls and orchestration steps.
Why Qloo matters
A generic language model can suggest popular venues or media. This application uses Qloo for the core entity-resolution and affinity-recommendation steps. Without Qloo, the application would lose the taste-graph recommendations, confirmed entity provenance and location-aware recommendations that define the product.
Architecture
- FastAPI backend with a server-side Qloo API client.
- Qloo
/searchfor entity resolution. - Qloo
/v2/insightsfor cultural affinity recommendations. - Responsive HTML/CSS/JavaScript frontend served by the same backend.
- Docker deployment configuration and synthetic unit tests.
Privacy and limitations
Only cultural references the user explicitly enters, selected Qloo entity IDs and the chosen city are sent to Qloo. The app does not send household memory, health notes, email addresses or personal identifiers, and it does not save taste profiles. The ambient scene is a preview preset; this prototype does not control physical devices. Qloo affinity is aggregate cultural data, not proof of an individual's preferences.
Build and validation status
The source code is now maintained in a dedicated public repository. Synthetic tests and CI are included. Live Qloo API validation and public hosting remain the final steps before submission.
Built With
- css
- docker
- fastapi
- html
- httpx
- javascript
- python
- qloo-api
- qloo-taste-graph

Log in or sign up for Devpost to join the conversation.