Build a semantic layer for agentic AI on AWS with Strategy Mosaic and Amazon Bedrock AgentCore
The third post in our semantic-layer series builds a customer 360 across Amazon Redshift and Amazon Aurora with Strategy Mosaic, exposes it to a Strands agent on Amazon Bedrock AgentCore through the Mosaic MCP server, and enforces per-user row and column security end-to-end.
First seen by Oriel on .
AI summary
The post shows how Strategy Mosaic links Redshift and Aurora customer data into a governed semantic layer that a Strands agent accesses through Amazon Bedrock AgentCore and MCP.
- Mosaic defines separate Purchase and Customer models, then links them on shared Customer and Time attributes without ETL or a materialized cross-engine join.
- For cross-source queries, Mosaic pushes native SQL to each source and combines intermediate results in its engine using declared model links.
- Attributes, hierarchies, and reusable metric and filter objects encode business meaning, including customer identity resolution, geographic rollups, and spend calculations.
- The agent connects to Mosaic through MCP using the human caller’s OAuth token; Mosaic checks certified-model access and applies row-level and column-level security.
- The walkthrough uses Mosaic Models and Strategy One (March 2026); the text also describes publishing governed definitions to other clients and choosing live or in-memory execution.
#semantic layers #agentic ai #amazon bedrock #model context protocol
Generated by gpt-5.6-sol. Check the original for details.
Aliases
| Source | External ID | URL |
|---|---|---|
| AWS Machine Learning Blog | 27b90480a1f289c553a54c263feab25503621afe | https://aws.amazon.com/blogs/machine-learning/build-a-semantic-layer-for-agentic-ai-on-aws-with-strategy-mosaic-and-amazon-bedrock-agentcore/ Read discussion |