QuickSight (now Amazon Quick Sight)
The name has changed
Section titled “The name has changed”Amazon QuickSight is now Amazon Quick Sight, a feature within Amazon Quick. AWS’s own wording is: “Amazon Quick evolved from Amazon QuickSight. QuickSight continues as Amazon Quick Sight, a feature within Quick. All existing QuickSight APIs, SDKs, and integrations continue to work without changes.”
The practical consequences:
- Nothing you have built breaks. The IAM action prefix is still
quicksight:, the ARNs are stillarn:aws:quicksight:..., and the APIs and SDKs are unchanged. - The surrounding product is larger. Amazon Quick now also covers Quick Flows (task automation), Quick Automate (agent-driven process automation), Quick Index (grounding AI answers in your own documents), Quick Research, and natural-language app building. Quick Sight is the analytics and dashboard piece of that.
- Documentation and console labelling now say Quick or Quick Sight. Search for both names when looking things up.
Everything below describes the analytics capability, whatever it is called this quarter.
What it is
Section titled “What it is”A fully managed, serverless business intelligence service for building interactive dashboards and reports and distributing them to people who are not analysts. It runs in a browser and in mobile apps, and dashboards can be embedded in your own applications.
Its position in the AWS estate is straightforward: it is the tool for putting business-facing charts in front of business-facing people. It is not an operational monitoring tool — for that, CloudWatch dashboards and Managed Grafana are the right layer.
SPICE — the Super-fast, Parallel, In-memory Calculation Engine — is the in-memory store behind the service. Importing a dataset into SPICE decouples dashboard performance from the source system: queries hit the cache rather than hammering the production database every time somebody opens a dashboard. The alternative is direct query, where each interaction goes to the source.
The trade-off is the usual one. SPICE is fast and protects the source, but the data is as fresh as the last ingestion. Direct query is always current and always costs the source system something.
Data sources
Section titled “Data sources”Native AWS sources include Amazon Athena, Aurora, RDS, Redshift, S3, OpenSearch Service, IoT Analytics and Timestream. Beyond those it connects to the common relational databases, to a range of third-party SaaS applications, and to anything reachable through AWS Glue.
One caveat on that list: Amazon Timestream for LiveAnalytics has been closed to new customers since 20 June 2025. Existing workloads are unaffected and AWS continues to invest in the service, but if you do not already have it, Amazon Timestream for InfluxDB is the alternative AWS recommends.
A dataset can combine several sources, which is what makes it useful for reporting that spans systems.
Reaching data that is not query-ready
Section titled “Reaching data that is not query-ready”The usual pattern for raw data sitting in S3 is to put a catalogue in front of it before the BI tool sees it:
- An AWS Glue crawler infers the schema of the raw data.
- The result lands in the Glue Data Catalog as table metadata.
- Athena can then query the S3 data directly using that catalogue, and Quick Sight uses Athena as its data source.
- Where the data needs reshaping first, Glue ETL jobs write the cleaned output to S3, Redshift or a Lake Formation-governed lake, and the dashboard reads from there.
The same catalogue serves EMR and Redshift Spectrum, so this is worth building once rather than per-tool.
Editions, and what is Enterprise-only
Section titled “Editions, and what is Enterprise-only”There were two editions, Standard and Enterprise, and a great deal of older
writing still compares them. A new account is now Enterprise. The
CreateAccountSubscription API still accepts STANDARD as a value, but its
own documentation says you can choose from ENTERPRISE or ENTERPRISE_AND_Q;
Standard survives only for accounts that already had it. Read “Enterprise
edition only” in the documentation as “available”, not as an upsell you have
to plan around.
The distinction still matters when you are reading the documentation, because the security and governance features are where the two diverged, and this is the point most often got wrong:
- Row-level security (RLS) — Enterprise edition only. It restricts a dataset by rows, driven by a rules dataset keyed on user or group name (or user and group ARN). It applies to string fields only; dates and numeric fields are not supported. Note that a user or group with no rule sees nothing at all, and rows whose restricted field is NULL or an empty string are skipped.
- Column-level security (CLS) — Enterprise edition.
- Groups — Enterprise edition. A legacy Standard account manages individual users only, which is why RLS rules there can only be keyed on user names.
Quick Sight users and groups are its own; they are managed inside the service and are distinct from IAM principals, though the service integrates with IAM Identity Center and with SAML identity providers.
Pricing is per user per month, across four roles: Author and Author Pro for people who build content, Reader and Reader Pro for people who consume it, with the Pro tiers adding the AI capabilities. Reader access can alternatively be bought as reader capacity — a bulk monthly or annual commitment of sessions, which is the cheaper shape when a large audience each opens a dashboard occasionally. Check the current pricing page before committing; the roles and rates have changed more than once.
Other capabilities
Section titled “Other capabilities”- Natural-language querying, and ML-driven anomaly detection and forecasting over your data.
- Scheduled email reports.
- Filters, parameters and custom actions for interactive dashboards.
- Embedded analytics, using temporary credentials to grant access to an embedded dashboard.