How to Build Your First Sigma Computing Workbook

Sigma Computing sits between a spreadsheet and a business intelligence tool. It connects directly to your cloud data warehouse and lets you explore, filter, and visualize data through a familiar row-and-column interface, without requiring SQL. A 2025 DataCamp course survey found that Sigma users reported up to 60 percent faster first-dashboard delivery compared to teams migrating from traditional BI tools that required dedicated data engineering support. This guide walks through the complete process: connecting a data source, building a workbook, adding charts, and publishing a live dashboard.

What Makes Sigma Different From Other BI Tools

Most business intelligence platforms either require SQL knowledge or store a copy of your data in a proprietary layer. Sigma does neither. It queries your cloud warehouse directly at runtime, which means your dashboards always reflect current data. The interface uses spreadsheet conventions: rows, columns, grouping, and formulas that look like Excel functions.

Unlike a spreadsheet, Sigma manages calculations at the column level. A formula applied to one cell in a column applies to every row in that column automatically. This architecture handles millions of rows without performance degradation, because the computation runs on the warehouse, not in the browser.

Sigma supports Snowflake, Google BigQuery, Databricks, Redshift, PostgreSQL, MySQL, and AlloyDB as data connections. For teams without a warehouse, Sigma also accepts direct CSV uploads.

What You Need Before Starting

To follow this tutorial, you need:

• A Sigma account (a free trial is available at sigmacomputing.com)
• Access to a cloud data warehouse, or a CSV file with sample data
• Account type set to Creator or higher (assigned by your Sigma administrator)

Sigma's permission model is hierarchical: organization level, then team level, then individual account type. The "Create, edit, and publish workbooks" permission is required to complete this guide. If you cannot create a workbook, ask your Sigma administrator to upgrade your account type.

Step 1: Connect Your Data Source

After logging in, navigate to Administration in the left sidebar. Select Connections, then click Add Connection.

Choose your warehouse type from the list. For Snowflake, you will enter your account identifier, warehouse name, database, schema, and the username and password for a service account with read access. Sigma recommends creating a dedicated Snowflake role for Sigma with SELECT privileges on the schemas you plan to query.

Once you enter the credentials, click Save. Sigma tests the connection automatically and shows a green status if the credentials are valid. The connection will now appear in the data panel whenever you create a workbook.

If you are using a CSV instead of a warehouse connection, skip the connection step. You will upload the file directly when creating your first workbook element in Step 3.

Step 2: Create a New Workbook

From the Sigma home page, click the Create New button (the plus icon in the upper left), then select Workbook. A new exploration opens in a blank canvas.

An exploration is an unsaved workbook. It persists in your home page for 30 days, but it is not shared with your team until you publish it. This gives you a safe space to experiment before committing to a final structure.

Workbooks are organized into pages, which function like tabs in a spreadsheet. Each page can hold multiple elements: data tables, pivot tables, charts, text blocks, images, and control elements like date pickers and dropdown filters. A single page can serve as a standalone dashboard, or you can use multiple pages to separate views by audience.

Step 3: Add a Data Table

Click Add Element in the canvas. Choose Table from the element type list.

In the data panel on the left, expand your connection and navigate to the table or view you want to analyze. Click the table name to load it into the workbook. Sigma pulls a live preview of the data directly from your warehouse.

You will see columns displayed as headers. To change the order, drag column headers left or right. To hide a column from the view without removing it from the underlying query, right-click the header and select Hide Column.

If you are using a CSV, click Upload CSV in the data panel instead, and select your file. Sigma parses the columns and data types automatically.

Step 4: Group and Aggregate Your Data

Raw tables show individual rows. To summarize data, use grouping.

Click the column you want to group by, for example a Region or Category column. In the column menu, select Group Column. Sigma collapses the rows into unique values for that column and displays an aggregation row for each group.

To add a metric, click the plus icon in the column header area and select Add Calculation. Type a formula like Sum([Revenue]) or Count([Order ID]). Sigma uses a formula syntax similar to Excel. The result appears as a new column showing the aggregate per group.

You can add multiple grouping columns to create hierarchical summaries. Clicking a group row expands it to show the underlying rows, which is useful for drilling into specific segments.

Step 5: Build a Chart

With your grouped table selected, click the chart icon in the element toolbar. Sigma reads the current table structure and suggests a chart type. For grouped data with a numeric metric, it typically defaults to a bar chart.

To change the chart type, open the Chart dropdown on the right panel and select from bar, line, scatter, pie, area, or other options. Assign your grouping column to the X axis and your metric to the Y axis using the axis pickers.

Titles, axis labels, and color palettes are editable in the Formatting section below the chart configuration. Click the legend to adjust series colors individually.

Charts in Sigma remain connected to the underlying data table. If you update the table's filters or grouping, the chart updates automatically.

Step 6: Add Filters and Controls

Filters let dashboard viewers narrow the data without editing the workbook. Click Add Element and select Control, then choose the control type: a date range picker, a dropdown list, or a text search box.

For a dropdown filter on a Region column, select List control. Point it at the Region column in your data element. Sigma populates the dropdown with the unique values from that column at query time.

To connect the control to your table and chart, click the control element, go to the control's settings panel, and select the elements it should filter. When a viewer selects a region from the dropdown, both the table and the chart update to show only that region's data.

Controls can be scoped to a single page or applied across all pages in the workbook.

Step 7: Publish and Share

When your workbook is ready, click Publish in the upper right corner. You will be prompted to name the workbook. After publishing, it appears in your Sigma home page under the folder you select.

To share the workbook, open the Share menu. You can invite specific users or teams by email, or generate a shareable link. Permissions cascade from the connection level: users can only see data they have warehouse access to, even when viewing a shared workbook.

For teams that need to share live dashboards with stakeholders who do not have Sigma accounts, Sigma supports embedded sharing through a signed URL, which renders the workbook inside another application or web page.

If your team is starting from raw files rather than a connected warehouse, VSLZ AI offers a similar workflow from a direct file upload, running analysis and charts from a plain-language prompt without any configuration.

Practical Notes

Sigma creates no copies of your data. Every query runs against your live warehouse, which keeps storage costs predictable. For large tables, applying filters early in the workbook reduces query time significantly: filter at the table level before adding aggregations, not after.

Unpublished explorations expire after 30 days. Publish early to avoid losing work. Saving frequently during exploration is not necessary since Sigma auto-saves explorations, but publishing is required to make a workbook visible to your team.