Build a semantic layer for agentic AI on AWS with Denodo and Amazon Bedrock AgentCore

This article demonstrates building a semantic layer for agentic AI using Denodo Platform and Amazon Bedrock AgentCore to enable governed access to distributed enterprise data across multiple sources.

• Federate Amazon Aurora, Redshift, and on-premises PostgreSQL into a unified customer 360 view queryable by agents
• Expose semantic views through Denodo MCP Server registered as AgentCore Gateway target for secure agent access
• Propagate user identity end-to-end from enterprise IdP through AgentCore to Denodo for consistent access control
• Apply Global Security Policies for row and column-level filtering based on caller roles and attributes
• Define reusable business metrics in metric views that agents reference without owning join logic
• Deploy Denodo as fully managed Agora service or self-managed on EC2/EKS within customer AWS account

The semantic layer enables agents to query live enterprise data across cloud and on-premises systems with governed access, consistent policies, and business-meaningful abstractions rather than raw schema.