SoniqueDNA

Beyond genres: discover music through your cultural affinities and interest

Comment 2

Project Story: SoniqueDNA

Inspiration

Traditional music discovery platforms rely solely on genre-based algorithms, creating echo chambers where users get stuck listening to the same types of music. We were inspired by the Qloo Global Hackathon's challenge to explore the intersection of LLMs and cultural intelligence. We wanted to solve the fundamental problem: how can we discover music that truly resonates with our lifestyle and cultural preferences, not just our existing musical taste?

The Problem We Solved

  • Genre Bubbles: Users trapped in repetitive recommendations based only on musical similarity
  • Lack of Context: No understanding of user's emotional state, lifestyle, or cultural interests
  • Shallow Discovery: Missing connections between music and broader cultural preferences
  • Privacy Concerns: Traditional platforms require extensive personal data collection

Our Solution: Cultural Intelligence

SoniqueDNA leverages Qloo's Taste AI™ API and Google's Gemini 2.0 Flash to create a revolutionary music discovery experience:

$$ \text{Recommendation} = f(\text{Spotify Data}, \text{Gemini Context}, \text{Qloo Cultural Intelligence}) $$

Technical Architecture

  1. Natural Language Processing: Users describe their mood/context in plain English
  2. Gemini AI Analysis: Context analysis generates cultural tags and emotional insights
  3. Spotify Integration: Gathers user's listening history and preferences
  4. Qloo Cultural Intelligence: Cross-domain recommendations through cultural affinities
  5. Unified Recommendation Engine: Combines all signals for personalized results

Key Features Built

  • Cross-Domain Discovery: Find artists through movies, books, TV shows, and cultural interests
  • Location-Based Intelligence: Cultural recommendations based on geographic context
  • Real-Time Chat Interface: Natural conversation with AI for music discovery
  • Comprehensive Analytics: Track cultural evolution and discovery patterns
  • Privacy-First Design: Cultural insights without requiring personal data

What We Learned

  • Cultural Intelligence: The power of connecting music to broader lifestyle preferences
  • LLM Integration: How to effectively combine multiple AI services for enhanced user experience
  • Cross-Domain APIs: Leveraging cultural data to create meaningful recommendations
  • User Experience: The importance of natural language interfaces in music discovery

Challenges Faced

  1. API Integration Complexity: Coordinating Spotify, Qloo, and Gemini APIs with different response formats
  2. Real-Time Processing: Managing multiple API calls while maintaining responsive user experience
  3. Cultural Tag Optimization: Finding the right balance between cultural relevance and musical quality
  4. Error Handling: Robust fallback systems when external APIs are unavailable
  5. Performance Optimization: Caching and batch processing for faster recommendations

Technical Stack

  • Frontend: React 18.3.1, TypeScript, Tailwind CSS, Shadcn/ui
  • Backend: Flask 2.3.3, Python, SQLite
  • AI Services: Google Gemini 2.0 Flash, Qloo Taste AI™ API
  • Music API: Spotify Web API
  • Deployment: Vite, React Query, Real-time progress tracking

Impact

SoniqueDNA represents a paradigm shift in music discovery:

  • Personalization: Understanding user context beyond musical preferences
  • Cultural Connection: Bridging music with broader cultural interests
  • Privacy Protection: Cultural insights without personal data collection
  • User Empowerment: Natural language control over discovery experience

Future Vision

We envision SoniqueDNA as the foundation for a new era of culturally intelligent entertainment discovery, where users can explore music, movies, books, and cultural content through unified, context-aware recommendations that truly understand their lifestyle and preferences.

Built with ❤️ for the Qloo Global Hackathon

Share this project:

Updates