Skip to content
PuglieseWeb Engineering HandbookSystemDox Pages

Business Intelligence

Amazon’s business intelligence product on AWS is Amazon Quick Suite. Until 9 October 2025 it was called Amazon QuickSight, and most existing architecture material still uses the old name.

Within Quick Suite, the BI capability — dashboards, analyses and datasets — is Amazon Quick Sight. The suite also includes Quick Research for cited answers over enterprise and public data, Quick Flows and Quick Automate for workflow automation, and Quick Index as a shared knowledge base over company documents and data, reached through a natural-language interface.

Quick Sight features

Section titled “Quick Sight features”
  • Serverless — no capacity to provision
  • Pay-per-session pricing for readers, so cost scales with use rather than with seat count
  • SPICE, the in-memory calculation engine, for fast queries over imported data
  • Interactive dashboards and visualisations, embeddable in your own applications
  • Direct connection to AWS data sources
  • Natural-language question answering through Amazon Q in Quick Suite. This superseded the earlier “QuickSight Q” feature: Amazon Q in QuickSight became generally available in April 2024 and the capability now sits inside Quick Suite.

Services that feed a BI pipeline

Section titled “Services that feed a BI pipeline”

Amazon Redshift — the warehouse most BI workloads read from. Columnar storage, massively parallel query execution, and connectors for most BI tools. Redshift Spectrum extends queries out to data still sitting in S3.

AWS Glue — the ETL and cataloguing layer: crawlers infer schema, the Data Catalog holds it, and jobs transform data into the shape reporting needs. See Athena and AWS Glue.

Amazon Athena — SQL over data in S3 with no loading step, for exploratory analysis and for dashboards over data that does not justify a warehouse.

AWS Lake Formation — permissions and governance over the data lake.

Amazon OpenSearch Service — search and log analytics, where the questions are about events rather than aggregates.

Amazon EMR — heavy transformation before the data reaches the warehouse.

Common shapes

Section titled “Common shapes”

The general pipeline:

Data sources -> ETL/processing -> storage -> analysis -> visualisation
(S3, RDS)       (Glue, Lambda)    (Redshift) (Athena)   (Quick Sight)

The warehouse-and-lake variant, where curated data lands in Redshift for modelled reporting while the lake stays queryable directly:

Raw data -> Glue ETL -> Redshift -> Quick Sight
   +------> S3 data lake <--> Athena ---^

Which one applies depends on query volume and predictability. Repeated dashboard refreshes over the same aggregates favour the warehouse; occasional exploration over raw data favours Athena against the lake.