PureWave: Culturally Grounded Taste Intelligence Agent

Inspiration

AI agents are rapidly mastering logical reasoning, code synthesis, and task automation. Yet when deployed into creative, cultural, and commercial industries, they remain profoundly culturally blind. Standard LLM agents rely on sterile acoustic statistics or generic web scrapers that fail to understand why audiences love what they love.

We built PureWave to solve this fundamental limitation by wiring Qloo's 250M+ Entity Taste Graph into our sovereign multi-agent architecture. By grounding autonomous music intelligence agents in real human cultural taste, PureWave transforms how independent artists and creative labels plan releases, sequence tracks, discover cross-domain fan demographics, and forge brand partnerships.

What It Does

• Cross-Domain Taste Grounding: Seamlessly maps musical artists (urn:entity:artist) against global taste clusters across streetwear fashion (urn:entity:brand), urban cultural destinations (urn:entity:destination), cult cinema (urn:entity:movie), and influential podcasting (urn:entity:podcast).
• Autonomous Release Strategy Synthesis: Takes artist seeds (such as Moski's 144 Raising EP) and queries Qloo's Insights API to compute statistical affinity overlaps, formulating precise PR angles and brand collaboration targets.
• Cultural Tour & Activation Routing: Identifies high-resonance geographic hubs (e.g. Shoreditch & Hackney in London) and venue typologies where an artist's specific audience profile naturally congregates.
• A&R Affinity Scoring: Uncovers non-obvious musical collaborations and feature opportunities by identifying shared cultural DNA between seemingly disparate artists.
• Zero-Leak IP Containment: Operates as a public verification harness and API gateway, ensuring that proprietary audio masters, private trading algorithms, and internal model weights remain strictly protected behind private infrastructure.

How We Built It

• Qloo Insights API Integration: Engineered runtime/qloo_client.py adhering to Qloo's modern /v2/insights schema with parameterised filter.type routing and robust fallback mechanisms.
• Autonomous PureWave Agent: Authored agents/purewave/taste_agent.py to orchestrate multi-domain cultural queries, affinity normalization, and actionable strategy generation.
• Web-MPC Cryptographic Coordination: Linked taste intelligence into the Sovereign Conglomerate's threshold signature system (crypto/threshold_mpc.py), ensuring authenticated execution across ephemeral container pods.
• High-Performance Deployment: Hosted live on Google Cloud Run (https://app.moski.ai) with sub-120ms response times and serverless scaling.
• Verified Open-Source Repository: Published full test suites and client harnesses on GitLab (https://gitlab.com/moski-labs/sovereign-ai-conglomerate) with automated CI/CD pipeline verification.

Challenges We Ran Into

• Multi-Domain Normalisation: Correlating heterogeneous taste dimensions (e.g., comparing relative affinity scores between technical outerwear fashion and cult film titles) required developing balanced scoring heuristics to avoid popularity bias.
• Low-Latency Agent Execution: Harmonising sequential Qloo Insights API queries across four distinct entity categories while preserving sub-second agentic decision loops.

Accomplishments That We're Proud Of

• Unlocking genuine cultural intelligence for autonomous agents, bridging the gap between raw machine reasoning and human taste.
• Achieving 100% automated test coverage across our Qloo client and PureWave reasoning pipeline.
• Designing an architecture that provides deep creative value to independent music creators while safeguarding proprietary intellectual property.

What We Learned

Qloo's Taste Graph proves that cultural affinity is structured, interconnected, and highly predictive. When agents are provided with true cultural grounding, their recommendations shift from generic platitudes to deeply resonant, actionable strategies that creators can execute immediately.

What's Next for PureWave

• Expanding real-time streaming listeners' taste graph clustering directly from purewave.ai playback sessions.
• Adding automated merchandise capsule design generation based on high-affinity fashion brands.
• Deploying autonomous agentic booking negotiations with high-affinity independent music venues across the UK and Europe.

Built With

• autonomous-agents
• cloud-run
• fast-api
• gitlab-ci
• multi-agent
• music-tech
• python
• qloo-api
• taste-graph