Amazon Quick generates dashboards from natural language prompts
Turn hours of dashboard building into minutes by simply describing what you want in plain English.
View original announcement →Visual Summary
graph TD
A{{Generate Analysis}}:::announced
B((BI Author)):::external
C([Natural Language Prompt]):::feature
D(Amazon QuickSight):::compute
E([Editable Plan]):::feature
F(Datasets):::storage
G([Auto Visualizations]):::feature
H([Calculated Fields]):::feature
I([Filter Controls]):::feature
J((Publishing/Embedding)):::external
B ==>|"describes goals"| C
C ==>|"inputs to"| A
F -->|"up to 3 datasets"| A
A -->|"presents"| E
E -->|"confirms"| D
D -->|"generates"| G
D -->|"creates"| H
D -->|"adds"| I
G -.->|"integrates"| J
H -.->|"integrates"| J
classDef announced fill:#ff9900,stroke:#ec7211,color:#fff,font-weight:bold
classDef compute fill:#e3f2fd,stroke:#1565c0,color:#1565c0
classDef storage fill:#e8f5e9,stroke:#2e7d32,color:#2e7d32
classDef feature fill:#fff3e0,stroke:#e65100,color:#e65100
classDef external fill:#f5f5f5,stroke:#616161,color:#616161
What's New
Amazon Quick has launched Generate Analysis, a generally available feature that enables users to create fully structured BI dashboards from natural language prompts. Users describe the dashboard they want in plain English, select up to three datasets, and Amazon Quick automatically generates organized sheets, appropriate visualizations, filter controls, and calculated fields such as year-over-year and month-over-month comparisons. This capability compresses what was previously hours of manual dashboard configuration into a process that takes minutes.
How It Works
- The workflow begins when a user opens a dataset in Amazon Quick and selects "Generate Analysis." The user provides a natural language description of their analytical goals (e.g., "create a sales performance dashboard with revenue trends, regional comparisons, and month-over-month growth") and selects up to three datasets to incorporate.
- Amazon Quick then presents an editable plan for the user to review and adjust before committing to generation — providing a human-in-the-loop checkpoint.
- Upon confirmation, the system produces a complete analysis containing multiple organized sheets, auto-selected chart types appropriate to the underlying data, interactive filter controls for dimensional exploration, and pre-built calculated fields for common time-series comparisons.
- The generated analysis is fully editable and integrates natively with existing Amazon Quick workflows, including publishing, embedding in applications, CI/CD pipelines, and standard point-and-click authoring.
Why It's Important
- Dashboard creation has historically been one of the most time-intensive tasks in BI workflows, requiring authors to manually configure visuals, define calculated fields, set up filters, and organize layout — often taking hours per dashboard.
- Generate Analysis dramatically lowers this barrier, enabling both experienced analysts and less technical business users to produce production-ready dashboards rapidly.
- The inclusion of an editable plan before generation is particularly significant, as it preserves author control and reduces the risk of AI-generated outputs that miss business intent.
- The seamless integration with existing publishing and embedding workflows means generated dashboards can move directly into production pipelines without rework, accelerating time-to-insight across organizations.
How It's Different
- Prior to Generate Analysis, Amazon Quick (formerly Amazon QuickSight) offered Q, a natural language query interface that answered individual ad-hoc questions and generated single visuals or simple analyses.
- Generate Analysis is a fundamentally different capability — it targets the creation of complete, multi-sheet dashboards rather than answering point-in-time questions.
- Where Q was optimized for exploratory querying, Generate Analysis is optimized for structured dashboard authoring at scale.
- Compared to manual dashboard creation, it eliminates the need to individually configure each visual, write calculated fields from scratch, or manually wire up filter controls.
- Relative to third-party AI-assisted BI tools, the tight integration with Amazon Quick's native data layer, publishing infrastructure, and embedding APIs provides a more cohesive end-to-end experience without requiring data to leave the AWS ecosystem.
When to Prefer It
- Generate Analysis is best suited for scenarios where speed of dashboard creation is a priority and the analytical requirements can be expressed in natural language with reasonable specificity.
- It is particularly valuable for rapid prototyping — generating a baseline dashboard that authors then refine — rather than replacing highly customized, pixel-perfect dashboard builds.
- Teams onboarding new datasets, responding to ad-hoc stakeholder requests for new dashboards, or standardizing reporting across business units will benefit most.
- It is also well-suited for organizations looking to democratize dashboard creation beyond a small group of expert BI developers.
- However, for dashboards requiring highly complex custom calculations, intricate layout requirements, or advanced interactivity beyond standard filters, manual authoring or iterative refinement of the generated output will still be necessary.
Availability
- Generate Analysis is generally available (GA) as of May 4, 2026, across all AWS Regions where Amazon Quick is available.
- At launch, the feature is accessible to users on the Enterprise subscription with Author Pro roles.
- Additionally, standard Authors on Amazon Quick Enterprise plans have promotional access through December 2026, provided their organization administrator has not restricted access to the feature.
- There is no indication of a Reader-tier or Standard subscription tier availability at this time.
- The three-dataset limit per generated analysis is a notable constraint for use cases requiring broader data integration.
- Users should consult the Amazon Quick User Guide section "Generating an analysis with natural language prompts" for detailed configuration guidance.
