In Amazon Quick, Analysis is the authoring surface where you build and arrange visuals, filters, and calculated fields across multiple sheets. When you’re ready to share, you publish the analysis as a dashboard. This new generative AI capability creates the analysis and you refine and publish it as a dashboard with a single click.

In this post, we walk through generating an analysis from a prompt, reviewing the plan, and exploring the completed output.

To generate an analysis, start by selecting the data that you want to analyze. In Amazon Quick, your data is stored in datasets, which connect to sources such as Amazon Redshift, Amazon Simple Storage Service (Amazon S3), or uploaded files. With your dataset ready, you describe what you want to see, review a plan, and generate.

Open a dataset in Amazon Quick and choose Generate analysis. You can also start from the Analyses page. Select 1–3 datasets for the analysis. If your data spans multiple tables—orders in one dataset and products in another—you can select them together.

Write a natural language prompt describing the insights you want to author in the Analysis. Describe the business questions, the metrics that you care about, and how you want the information organized across sheets. For example: “Create an operations dashboard showing order volume trends, revenue KPIs, delivery performance comparing estimated vs actual delivery dates, and product category breakdown by revenue and order count. Include calculated fields for total revenue, average order value, and month-over-month order growth.”