Amazon Web Services now introduces Context Ontology Accelerator

Today, AWS announces Context Ontology Accelerator, an open source accelerator that helps organizations build an ontology of their business — a machine-readable model of their products, customers, policies, and the rules that define how they operate so AI agents make accurate, consistent, explainable, and auditable decisions. Context Ontology Accelerator connects to your structured and/or unstructured data sources, and uses AI to draft an ontology; your domain experts review, edit, and approve every element of it. The approved ontology is stored in a knowledge graph you own, expressed in open W3C standards so it works with any standards-based tooling, and any agent can consume it through the included Model Context Protocol (MCP) server.

Organizations building AI agents need more trust in their agents' decisions, with outputs they can explain and audit. The context those agents require — entities, rules, policies, relationships — is scattered across dozens of systems with conflicting schemas and naming conventions. Teams spend months reconciling definitions and encoding them into prompts and point integrations, and because the results are not traceable, teams block agents from moving from proof-of-concept to production. Context Ontology Accelerator reduces what would take months of manual ontology authoring into days — from connecting data to serving governed answers to agents.

Context Ontology Accelerator is available today on GitHub under Apache 2.0. The initial release uses Amazon Neptune as the graph store, Amazon OpenSearch Serverless as the vector store, and foundation models hosted on Amazon Bedrock; deploy it yourself or engage AWS Professional Services. Context Ontology Accelerator’s managed, user-defined ontology capability will become a fully managed feature native to AWS Context. Customers who get started with Context Ontology Accelerator to create ontologies will be able to use and manage them with AWS Context. To learn more, read the documentation.