Generative AI created an extraction problem. It consumed the creative output of millions to build value that accrues to a few. The missing infrastructure is not better models, it is attribution. Who contributed what? Who refined it? Who owns the inference?
Esther Anglade will present a production system built on Neo4j that treats the knowledge graph not as a database, but as a ledger of intellectual contribution. In this architecture, every decision, constraint, preference, and skill node carries provenance: which agent or human created it, at what confidence, from which exchange, and when. This is not audit logging , it is the foundation for a new asset class.
You will see how graph-native memory solves three problems simultaneously. First, agent memory becomes reliable because it is accountable, agents earn trust through verifiable contribution history stored as graph properties. Second, costs collapse through stacking , instead of regenerating context from scratch, agents inherit refined knowledge from a shared graph, eliminating redundant inference. Third, security becomes architectural, ABAC at the node level, relationship-based access control, and cryptographic provenance make privacy a structural property, not a policy layer.
You will also learn why this matters beyond engineering. When contribution is traceable, it becomes compensable. Authors, researchers, and creators whose work informs AI systems gain a mechanism to prove and claim value. The graph becomes infrastructure for new creative rights, and the first step toward an economy where intelligence is owned by those who generate it.
Speaker
Esther Anglade
Researcher & Founder, OHACO Labs
Esther Anglade is the Founder of OHACO, where she is building attribution infrastructure for AI systems that make intellectual contribution visible, traceable, and ownable. Her production platform coordinates multiple AI agents across a live knowledge graph of over four million nodes, where every inference carries provenance and every agent earns trust through verifiable history. Her work synthesizes peer-reviewed research from ACL, EMNLP, CHI, USENIX Security, and NeurIPS into a practical architecture for sovereign AI memory. She believes the next economy will be built on graphs.
