The Personal Knowledge Vault: Art Collection Use Case

Session Track: Knowledge Graphs

Session Time: 08:00 pm - 08:30 pm November 6

Session description

Last year, Mike Morley introduced the Personal Knowledge Vault (PKV) as a system to support memory augmentation for individuals with functional memory challenges, using AI to make personal documents, notes, and media easily searchable and contextual. This year, he will demonstrate how that foundation has been extended to serve a not-for-profit art collection—digitizing, organizing, and retrieving historical records and the art, some dating back to 1883. You will learn how scanned documents, ranging from fountain pen ledgers to modern emails, are transformed into structured graph documents using NLP, NER, and asynchronous Celery processing. These are indexed in a Neo4j-powered GraphRAG system and linked to object records stored in a MinIO-compatible image store. QR codes allow instant access in galleries, while image-based recall and AI-transcribed stories further enrich interaction. Mike will also preview upcoming planned capabilities like synthetic memory generation using GenAI prompts and integration with OpenAI's MCP layer. This session will equip you with an understanding of how to design graph-based, agentic AI systems that capture, contextualize, and reanimate historical and object-linked knowledge.