Amazon Quick Introduces Agentic Catalog Experience for Enhanced Data Discovery and AI Integration

Amazon Quick has announced the Agentic Catalog Experience, an innovative AI-powered workflow aimed at bridging the last mile in enterprise data catalog utilization. This new feature enables data curators to efficiently discover relevant catalog assets using natural language, automatically create datasets and topics with inherited semantics, and deliver trusted, production-ready AI answers and dashboards.

The challenge addressed by this development is the disconnect between rich upstream metadata—such as descriptions, relationships, and governance details—and their manual recreation or manual discovery by end users. Despite extensive investments in catalog platforms like AWS Glue Data Catalog and Databricks Unity Catalog, organizations face difficulties in translating catalog metadata into actionable insights, leading to delays and semantic drift.

The Agentic Catalog Experience leverages the Quick Agent, which summarizes catalog contents, engages in natural language conversations, and surfaces relevant assets based on user queries. It can automatically create catalog representations—Datasets and Topics—with inherited semantics, including descriptions, relationships, and schema details, directly from upstream catalogs. This process reduces setup time from weeks to minutes and ensures that metadata remains synchronized with upstream sources.

Supported catalog platforms include AWS Glue Data Catalog and Databricks Unity Catalog, with plans to extend support to additional systems. The inherited metadata focuses on table and column definitions, primary and foreign key relationships, and schema models, providing a grounded context for end-user queries. The experience allows users to ask natural language questions, build dashboards, and share insights seamlessly, all grounded in the trusted catalog metadata.

The architecture emphasizes a consumer model, where Amazon Quick consumes metadata without duplicating data, maintaining the integrity and authority of upstream catalogs. Users can manually refresh inherited metadata to keep assets current, with scheduled automatic sync on the roadmap.

This enhancement enables organizations to deliver faster, more accurate insights, grounded in their existing data governance and semantic definitions, thereby improving trust and efficiency in enterprise analytics.