Exploring data

Introduction

With the data collected and prepared, it's time to start learning from it. In this guide you'll look inside a dataset, ask Autopilot questions in plain language, and generate your first chart — no formulas, no query building.

This guide continues from Prepping data.

Preview any dataset

The quickest way to see what's inside a dataset is to double-click its node on the canvas. A preview panel opens showing the dataset's columns and a sample of its rows, along with a Download button if you want a copy.

Try it: double-click the Superstore dataset node. You'll see the sales data exactly as it was collected.

Click Close preview in the top-right corner of the panel when you're done looking.

Ask questions in plain language

You don't need to scan thousands of rows to answer a question — just ask Autopilot. In the message box, type:

Which retailer had the most sales?

Autopilot analyzes the prepared data and answers directly in the chat.

You can keep the conversation going with follow-ups — compare regions, look at specific time periods, or drill into a category. Autopilot works with the latest version of your data based on the most recent workflow refresh pipeline, including the retailer names you added in the previous guide.

Generate a chart

Some answers are easier to see than to read. Ask for a visualization:

Show me total sales by retailer as a chart

Autopilot builds the chart and shares it right in the conversation.

What you built

You can now look inside any dataset with a double-click, get answers by asking questions in plain words, and turn those answers into charts. Next, turn this analysis into something you can share: continue to Building a report.