-
The agent workflow: intent → taste signals → Qloo → personalized recommendations.
-
TastePilot understands intent and turns preferences into personalized taste signals.
-
TastePilot AI — Personalized discovery powered by an AI taste agent.
Inspiration
People have more choices than ever, but discovering something that truly matches their personal taste is still difficult. A recommendation for a movie, restaurant, song, or activity often treats each choice separately instead of understanding the person behind those choices.
We wanted to explore whether an AI agent could understand a person's taste and turn it into a more meaningful discovery experience.
What We Are Building
TastePilot AI is an AI-powered taste agent designed to help people discover experiences that match their unique preferences.
Instead of simply asking for a category such as "recommend me a movie," users can describe what they like, how they feel, and what kind of experience they want. The agent uses Qloo's cultural intelligence to connect preferences across categories and generate personalized recommendations.
The goal is to create recommendations that feel connected rather than random — helping users discover combinations of movies, music, food, places, and experiences that make sense together.
Why Qloo
Qloo is central to our approach because the project is about understanding relationships between tastes and cultural preferences.
Rather than using Qloo as a simple data lookup, TastePilot AI is designed around its taste intelligence. The agent interprets the user's intent, identifies relevant preferences, uses Qloo to discover culturally compatible recommendations, and then turns those results into a useful response.
What We Hope to Learn
This hackathon is an opportunity for us to learn how to build a real-world AI agent around cultural intelligence, design a useful user experience, and turn an idea into a working product.
Challenges
The biggest challenge will be making recommendations feel genuinely personalized instead of generic. We also want the agent to explain its choices clearly and create useful connections between different types of experiences.
Our goal is to build a simple, polished prototype that demonstrates how taste-aware AI can make discovery more personal and useful.
Log in or sign up for Devpost to join the conversation.