Amazon Quick Sight: Business Intelligence & Dashboards

Quick is the BI consumption layer

Amazon Quick Sight is AWS's business-intelligence capability within Amazon Quick. AWS's current analytics decision guide refers to Quick at the platform level, while Quick Sight remains the BI product name and the QuickSight name still appears in APIs and certification material. The BI role is the important constant: authors connect data sources, create datasets and analyses, build visuals, and publish dashboards that readers can explore. Quick Sight is therefore normally downstream of storage and analytical processing rather than the service that performs upstream data engineering.

A data source defines connectivity to an upstream system, while a dataset represents the data Quick uses for analysis. An analysis is the authoring workspace where sheets and visuals are assembled. A dashboard is the published, read-oriented artifact derived from an analysis. Keeping these objects distinct helps explain permissions and refresh behavior: changing a source or dataset is different from publishing a new dashboard experience.

SPICE and direct query trade freshness for cached analytical performance

Quick can work with data through direct-query access where supported or import datasets into SPICE, its in-memory calculation engine. With direct query, visual interactions issue queries against the underlying data source, so source performance and availability influence the user experience. With SPICE, data is ingested into a managed in-memory analytical layer, decoupling dashboard interactions from repeated live-source queries.

SPICE datasets must be refreshed when upstream data changes and the dashboard needs the newer data. The architectural tradeoff is therefore freshness and source coupling versus cached interactive performance and reduced repeated source load. The correct mode depends on how quickly data must appear, how expensive live queries are, and whether the upstream source can sustain BI concurrency.

• Data source: connection to the upstream system.
• Dataset: selected and prepared data exposed to analysis.
• Analysis: authoring workspace for visuals and sheets.
• Dashboard: published experience for consumers.
• SPICE: imported in-memory data; direct query: query the source at interaction time.

Dashboards can sit on top of multiple analytics architectures

Quick can consume outputs from analytical services and data sources rather than forcing a single backend architecture. A team might query a Redshift warehouse for curated business metrics, use Athena-backed data-lake datasets for exploratory information, or ingest data into SPICE to make recurring dashboard exploration more responsive. The BI layer should not dictate where raw data is stored if another architecture better satisfies governance and processing requirements.

Embedding capabilities can place supported Quick experiences inside applications, while access controls determine which users and groups can reach assets. For sensitive BI, security must be designed at both layers: Quick permissions govern BI resources, while the underlying source and network configuration govern data access. Row- or column-level restrictions, where configured, should align with the business identity model rather than relying on dashboard layout to hide sensitive data.