AWS Launches 'AWS Context' to Build Organization-Wide AI Knowledge Graphs at AWS Summit NYC
Summary
- • AWS Context auto-maps enterprise data from lakes, warehouses, and databases into a shared, governed knowledge graph agents can query at runtime
- • AI agents access governed data relationships, business rules, and domain knowledge without custom ingestion pipelines or retrieval tuning
- • Extends Amazon Quick's personal knowledge graph — already processing millions of requests per day — to full organizational scale
- • No infrastructure to provision; exports to S3 in Apache Iceberg format; integrates with Glue Data Catalog, SageMaker Unified Studio, and Lake Formation
Details
AWS Context Announced at AWS Summit NYC
New AWS service automatically maps enterprise data relationships into a shared knowledge graph that AI agents can query at runtime for governed context — no pipeline building required
Extends Quick's Graph from Personal to Organizational
Amazon Quick's personal knowledge graph already processes millions of requests per day; AWS Context opens that same technology to all organizational agents and applications
Agentic Search with Governed Data Permissions
AI agents access the knowledge graph through agentic search; data stewards control which relationships are promoted to production and what domain knowledge and business rules are attached
Integrates with AWS Data Stack + Open Standards
Connects to Glue Data Catalog, SageMaker Unified Studio, and Lake Formation for governance; exports key context to Amazon S3 in Apache Iceberg format for tool-agnostic portability
Context Improves Continuously as Agents Use It
The graph observes which sources and join paths produce correct agent results, sharing that learning across all organizational agents — a compounding intelligence advantage that grows with usage
Key capabilities and context for AWS Context enterprise knowledge graph service
What This Means
AWS Context addresses a fundamental bottleneck in enterprise agentic AI: agents can only be as trustworthy as the context they can access, and most enterprise data is fragmented across dozens of disconnected systems. By creating a shared, governed knowledge graph that all agents in an organization can query at runtime, AWS is building a critical piece of the enterprise AI infrastructure stack. The self-improving nature of the graph — where one agent's successful query teaches all others — creates a compounding intelligence advantage over time. This also deepens the strategic value of the broader AWS ecosystem, as organizations adopting AWS Context become increasingly embedded in Amazon Quick, SageMaker, and Lake Formation.