Amazon QuickSight's Dataset Enrichment kills legacy sync headaches
Business context now lives in datasets, not separate topics—no more syncing.
Amazon QuickSight has unveiled Dataset Enrichment, a new data prep experience that moves business context—column descriptions, synonyms, calculated fields, custom instructions, and business rules—directly into the dataset itself. Previously, this metadata lived in separate legacy Topics objects that had to be manually synchronized with datasets, leading to drift in synonyms, diverging calculated fields, and silent breakage when datasets changed. With Dataset Enrichment, permissions, lineage, and governance are consolidated into a single asset, automatically inherited by anything built on top of it.
The migration from legacy Topics to Dataset Enrichment is not a cosmetic change; it establishes a forward-looking architecture supporting both deterministic BI workflows and flexible AI-driven analytics from a shared semantic foundation. Legacy Topics are being re-purposed as a multi-dataset semantic and reasoning layer, enabling cross-dataset Q&A and AI queries across multiple enriched datasets in one conversation. The post provides three migration scenarios with step-by-step guidance. Key unchanged aspects include rule datasets, SPICE storage, Direct Query, dashboards, and the Q&A interaction model. Dataset Enrichment sets the stage for catalog integration and a cleaner governance model.
- Column synonyms now stored in Additional Notes on the column instead of a separate legacy Topic object.
- Permissions and lineage simplified to one asset, eliminating two-asset sync and audit overhead.
- Legacy Topics re-purposed as a cross-dataset semantic layer for multi-dataset natural language and AI queries.
Why It Matters
Eliminates synchronization overhead and governance complexity, enabling cleaner AI-driven analytics across datasets.