AWS announced the integration of EDB Postgres AI with Amazon Bedrock AgentCore to build a semantic layer for agentic AI. This allows agents to query the layer to answer customer 360 questions. The integration is part of a series of posts showcasing different approaches to building a semantic layer for agentic AI.
What happened
AWS has announced the integration of EDB Postgres AI with Amazon Bedrock AgentCore to build a semantic layer for agentic AI. This integration enables agents to query the semantic layer to answer customer 360 (C360) questions. The semantic layer is built inside the operational Postgres database, with the schema and business definitions indexed together. Postgres GRANT, row-level security, and column-level grants enforce access control. The integration is part of a series of posts that showcase different approaches to building a semantic layer for agentic AI, including previous examples with Stardog’s knowledge graph, Denodo’s data virtualization platform, AtScale’s dimensional model, and Strategy Mosaic’s linked models.
FAQ
What is the use case for this integration?
The integration allows agents to query the semantic layer to answer customer 360 (C360) questions.
How does access control work?
Postgres GRANT, row-level security, and column-level grants enforce access control.
What is captured in OpenTelemetry traces?
AIDB captures OpenTelemetry traces for every SQL execution as rows in the aidb_otel schema, queryable with ordinary SQL.
Source: Amazon Web Services