Amazon Quick has introduced Live Data in Apps, which AWS says lets AI-built applications query governed Quick Sight datasets in real time rather than rely on static, build-time snapshots. Each query runs as the person viewing the application, applying that user’s existing row-level and column-level security permissions. The change connects two concerns for cloud architects: keeping application data current and controlling what each reader can see. AWS’s feature explanation describes both capabilities.
The distinction between querying a live dataset and accessing freshly updated underlying data matters, however. The supplied reporting describes different freshness behavior for SPICE and Direct Query datasets, alongside restrictions on access and combining data sources. Those details qualify what “live” means in practice.
Queries follow the viewer’s permissions
According to AWS, application users see data under their own Amazon Quick permissions, with previously configured row-level and column-level security applied. At the level described in the announcement, the mechanism is straightforward: the viewer’s identity governs the query and the data it can return. Row-level controls concern which records a reader can access; column-level controls concern which fields are accessible. AWS’s launch announcement says both carry through to the application.
The architectural implication is that sharing an application does not imply an identical data view for every reader. Where permissions differ, the same application can return different permitted data. That is a consequence of the described security model, rather than evidence of a separate application-specific policy system.
For architects, this makes the existing dataset permissions central to reviewing an app. The feature applies the controls already established; AWS’s description does not establish that those controls are correctly configured for a particular deployment.
Data freshness depends on the dataset mode
AWS says the capability lets application KPIs, charts, tables, and insights reflect the latest dataset data. A separate AIToday report says published apps query datasets on every open and support both SPICE, described as in-memory, and Direct Query datasets.
The report identifies an important difference:
| Dataset mode | Reported update behavior |
|---|---|
| SPICE | Updated data becomes visible after a dataset refresh. |
| Direct Query | No dataset refresh is required. |
These implementation details come from AIToday and remain uncorroborated by the AWS excerpts supplied for this article. They should therefore be treated as reported behavior, rather than independently tested results.
On that account, querying a SPICE dataset when an app opens does not remove the refresh dependency. The app can retrieve the dataset’s current contents while those contents still depend on the last refresh. For an architect assessing freshness, the useful question is therefore both when the application queries and when the selected dataset receives updated data. The evidence supplies no measured end-to-end latency or freshness guarantee.
Building the app and defining its audience
AWS describes a plain-language construction workflow: give the app builder chat agent the dataset name and describe the application to build. The company also positions these applications for business operations, including handling requests and approvals and triggering subsequent steps. Those are vendor-described uses; the supplied evidence does not detail their action mechanisms or demonstrate a completed deployment.
AIToday reports several prerequisites and constraints:
- Builders must be at least Reader Pro (Professional) in Amazon Quick.
- Viewers must be authenticated Quick users; anonymous or public access is unsupported.
- Each viewer must consent for each dataset on first use.
- Direct Query datasets from different sources cannot be used in the same app.
- Builders encounter an initial row limit during retrieval, although the supplied evidence gives no numerical limit.
These restrictions also rest on that single report and remain unconfirmed by the supplied AWS excerpts. They are material checks for a deployment review: authentication affects the intended audience, consent affects first use, and the Direct Query restriction affects dataset selection. The unspecified retrieval limit leaves an additional sizing question open.
AWS says the feature is available in all regions supported by apps in Amazon Quick; the supplied announcement does not enumerate them.
Live Data in Apps brings runtime querying and viewer-specific dataset permissions into AI-built applications, according to AWS. Its practical value depends on the selected dataset’s update behavior and the permissions already in place. For cloud architects, those are the concrete boundaries to evaluate before treating a generated app as a current, appropriately scoped view of business data.
