AWS has introduced new capabilities to help AI agents access and use enterprise data more effectively, enabling them to make trusted decisions by leveraging a shared, governed context layer. The company announced these innovations during its AWS Summit New York City keynote, including AWS Context, a service that maps data relationships into a knowledge graph, and updates to AWS Glue Data Catalog for better business context and semantic search. These tools are designed to help organizations govern data access and improve how AI agents reason over complex data environments. The new features also support integration with third-party systems and provide tools for developers to build and manage AI agents that can work with data more efficiently. Source: awsml
AWS Context automatically maps relationships across existing data into a knowledge graph, allowing AI agents to access governed data relationships, business rules, and domain knowledge at runtime. Data stewards can manage the graph through an intuitive console, reviewing inferred relationships, promoting them to production, and attaching domain-specific knowledge like business definitions and usage rules. This service extends the same knowledge graph technology that powers Amazon Quick, which already processes millions of requests per day. With AWS Context, users can create an organizational knowledge graph that agents and applications can draw from, enhancing decision-making through shared, structured data. Source: awsml
The AWS Glue Data Catalog now includes business context and semantic search features, allowing users to enrich data with business descriptions, glossary terms, and custom metadata. This makes it easier for humans and AI agents to discover and understand data by business meaning rather than technical metadata. Skill assets, a new asset type, can be associated with data assets to provide additional context and instructions for agents, reducing the need to re-teach every agent individually. These updates aim to help AI agents work more effectively with data by providing them with the necessary context and guidance. Source: awsml
