Agentic AI
AWS Context turns enterprise data into a knowledge graph for AI agents
AWS introduced a new service that automatically maps data relationships into knowledge graphs, enabling AI agents to access governed business context and domain knowledge at runtime.
The New Stack
The Problem with Raw Data#
AI systems need more than just volume—they need context. AWS recognizes that while its customers have vast pools of data scattered across data lakes, warehouses, and databases, that information often remains raw, unstructured, and inaccessible to AI agents. Much of it sits in undocumented institutional knowledge that agentic systems rarely touch.
Introducing AWS Context#
At its New York Summit, AWS unveiled AWS Context, a service designed to automatically map relationships across existing enterprise data into a knowledge graph. The service enables AI agents to access governed data relationships, business rules, and domain knowledge at runtime through agentic search capabilities.
Mai-Lan Tomsen Bukovec, AWS VP of technology (data and analytics), described it as a "data lake of nuance and information that AI agents swim in" to reason correctly and make sound business decisions. The comparison draws on how humans operate: contextual information informs better decision-making.
How Knowledge Graphs Work#
Knowledge graphs require more than keyword matching. They need structural and semantic traversal—the ability to hop across multiple information silos and aggregate context. This allows agents to explain complex relationships: why cybersecurity vulnerability A affects system compromise B, which depends on codebase C executing in application D, potentially affecting users X, Y, and Z.
Control and Governance#
Developers can exclude certain datasets (test data, sandbox environments) from AWS Context. The service continuously updates as data relationships change, keeping agents current without manual intervention. Each query is identity-aware, inheriting the calling user's IAM and Lake Formation permissions, ensuring agents only access authorized data. Every interaction is auditable.
AWS Context extends technology already powering Amazon Quick, the company's AI work assistant that connects scattered work across Slack, Teams, Outlook, CRMs, databases, and documents.
