Understand this story
Matthew Wallace, co-founder and CTO of Kamiwaza AI, argues that when enterprises build business ontologies for AI agents, they should ensure they are portable and avoid binding business processes to platforms they cannot leave. He recommends first distinguishing between the ontology (which defines entity types, the meaning of relationships, and rules) and the knowledge graph (which asserts specific entity relationships), and warns against vendors that make their own entities "first-class citizens" while restricting custom types. The article also notes that graph embeddings can efficiently retrieve relevant entities in large-scale knowledge graphs and rank candidate missing relationships, and that ontology building should proceed in stages, from automated discovery to human validation to continuous updating.
AI-generated from public reports
Latest development · Report
Portable Business Ontologies for AI Agents: How to Avoid Vendor Lock-inDevelopments
1 confirmed developments
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 3
Unite.AI
Portable Business Ontologies for AI Agents: How to Avoid Vendor Lock-in
Matthew Wallace, co-founder and CTO of Kamiwaza AI, argues that when enterprises build business ontologies for AI agents, they should ensure they are portable and avoid binding business processes to platforms they cannot leave. He recommends first distinguishing between an ontology (which defines entity types, the meaning of relationships, and rules) and a knowledge graph (which asserts specific entity relationships), and warns against vendors that make their own entities "first-class citizens" while restricting custom types. The article also notes that graph embeddings can efficiently retrieve relevant entities in large-scale knowledge graphs and rank candidate missing relationships, while ontology building should proceed in stages, from automated discovery to human validation to continuous updates.