diff --git a/site/sigmaguides/src/Fundamentals 1: Getting Around - v3/Fundamentals 1: Getting Around v3.md b/site/sigmaguides/src/Fundamentals 1: Getting Around - v3/Fundamentals 1: Getting Around v3.md index 0eb122ad..52db7008 100644 --- a/site/sigmaguides/src/Fundamentals 1: Getting Around - v3/Fundamentals 1: Getting Around v3.md +++ b/site/sigmaguides/src/Fundamentals 1: Getting Around - v3/Fundamentals 1: Getting Around v3.md @@ -6,10 +6,9 @@ status: Published feedback link: https://github.com/sigmacomputing/sigmaquickstarts/issues tags: default authors: phil@sigmacomputing.com -lastUpdated: 2026-06-20 +lastUpdated: 2026-07-02 # Fundamentals 01: Overview - ## Overview Duration: 6 @@ -43,7 +42,7 @@ If something isn’t working as expected, here’s how to [contact Sigma support ### About Sigma -Sigma is a front-end for your cloud data warehouse—an analytics platform for spreadsheets, interactive AI-enabled apps, and more. It centralizes your data, eliminates silos, and replaces fragmented tools for various use cases within a governed environment. Additionally, Sigma enables businesses to monetize their data through embedding Sigma in bespoke Saas applications. +Sigma is a front-end for your cloud data warehouse—an analytics platform for spreadsheets, interactive AI-enabled apps, and more. It centralizes your data, eliminates silos, and replaces fragmented tools for various use cases within a governed environment. Additionally, Sigma enables businesses to monetize their data through embedding Sigma in bespoke SaaS applications. Many solutions exist, but Sigma stands out. Here's why: @@ -57,7 +56,7 @@ For developers embedding Sigma in a larger platform, every feature is available With full feature parity and no coding required, Sigma is a strong choice for teams building embedded analytics. -Many customers are already using Sigma to replace existing, purpose built SaaS applications. +Many customers are already using Sigma to replace existing, purpose-built SaaS applications. ### Target Audience This QuickStart is designed for new Sigma users across three roles: @@ -118,7 +117,7 @@ After completing this QuickStart, here are suggested paths based on your role: - How to set up a Sigma trial and navigate the user interface - Key Sigma terminology and how workbooks, pages, and elements relate to each other - How to create, save, and share a workbook -- How to use Sigma Assistant for AI-powered analysis +- How to use Sigma Assistant for AI-powered analysis, and how it can plan and build dashboards and apps for you - How to organize content with pages, containers, modals, and other layout elements - How account types and scheduled exports work @@ -150,7 +149,7 @@ In your email application, open the email from Sigma and click `Verify Email Add NOTE:
If you have not received this email within 10 minutes, please check your spam folder, and if not there, ask for assistance from support. The email will be "from" Sigma Computing using the address: info@send.sigmacomputing.com. -When the page loads in your default browsers, enter in your `Company Name` and `Company Login URL` (if you prefer something different) +When the page loads in your default browser, enter in your `Company Name` and `Company Login URL` (if you prefer something different) @@ -208,7 +207,7 @@ The main areas are numbered and described below: **Notifications:** The bell icon shows alerts and activity relevant to your account. - **User Profile:** Click your initials in the upper right to access your profile, preferences, and sign-out and other options. + **User Profile:** Click your initials in the upper right to access your profile, preferences, sign-out, and other options. **Left sidebar** @@ -247,7 +246,7 @@ The main areas are numbered and described below: **Top Viewed:** As your team creates and uses content, the most-accessed workbooks will appear here. It will be empty at the start as there has been no activity yet. - **Invite your colleagues:** Provides a quick method for adding users to your account. This is a configurable convenience; users in Sigma can be managed in the administration section or externally via any SAML 2.0 compatible identity management provider (e.g., Okta, Auth0, etc.) + **Invite your colleagues:** Provides a quick method for adding users to your account. This is a configurable convenience; users in Sigma can be managed in the administration section or externally via any SAML 2.0 compatible identity management provider (e.g., Okta, Auth0) **Help:** The help button in the lower right corner provides access to documentation, release notes, and live chat support. @@ -369,17 +368,18 @@ Once configured, `Assistant` becomes available on the homepage: ### AI features in Sigma -Once a provider is configured, AI shows up in several places across the product: +Once a provider is configured, AI appears across the product — and it's designed as a deliberate spectrum, from helping with a single formula to bringing Sigma's data into the AI tools you use elsewhere. This QuickStart walks through them in that order, simplest to most advanced: +Two more AI touches show up while you work, though we don't walk through them here: Explain this Chart generates an AI description of any chart, and AI Queries run AI analysis on text data — sentiment, classification, summarization — as part of a workbook calculation. + @@ -407,6 +407,30 @@ There is also an additional layer worth knowing about: the **Sigma MCP Server**. ![Footer](assets/sigma_footer.png) +## Formula Assistant +Duration: 3 + +The simplest place AI shows up is the formula bar. The Formula Assistant lives in any workbook or data model and helps with a single calculation three ways: + +- **Write a formula** — describe what you want to calculate, and AI suggests the formula +- **Correct a formula** — when Sigma detects an error, AI interprets your intent and suggests a fix +- **Explain a formula** — AI describes, in plain language, what an existing formula does + +It becomes available when you select a column and open the formula bar. Click the AI wand icon to choose an option: + + + +For example, selecting a calculated column and choosing `Explain this formula` returns a plain-language description — the columns it references, how the calculation works, and what the result represents: + + + +This is the whole idea in miniature: AI does the mechanical work while you stay in control — reading, verifying, and adjusting what it produces. + +For more information, see [Use AI with formulas](https://help.sigmacomputing.com/docs/use-ai-with-formulas) + +![Footer](assets/sigma_footer.png) + + ## Sigma Assistant Duration: 8 @@ -481,10 +505,41 @@ Once in the workbook, we see that two tables and the chart are included (after m Now we can save it, refine it, or use Sigma's [drill-down](https://help.sigmacomputing.com/docs/view-underlying-data) capabilities for deeper analysis. We'll continue building on this workbook in the next section. - + + +![Footer](assets/sigma_footer.png) + + +## Building with Assistant +Duration: 4 + +The previous section showed Assistant answering a question and returning results you open in a workbook. Inside a workbook draft, Assistant does more than analyze — it can plan and build for you. Using natural language, it designs and constructs dashboards and AI apps: tables, charts, KPIs, controls, filters, input tables, and even the actions that make an app interactive. + +When editing a workbook draft, Assistant works in two additional modes: + +- **Plan mode** — Assistant gathers context and proposes a structured approach — recommended pages, elements, layout, and the data sources needed — without changing anything yet. Scoping in `Plan` mode first also saves time and AI credits, since you settle the structure before Assistant builds anything. +- **Build mode** — Assistant creates and modifies workbook content: elements, formatting, layout, navigation, filters, and actions. + +You can switch between modes at any time, and Assistant keeps the conversation's context as you move from scoping to building. + +Assistant is available from several entry points while editing a workbook draft: the empty-state prompt bar on a new page, the `Assistant panel` in the workbook header, the `⌘ + K` / `Ctrl + K` keyboard shortcut, and the `Ask or edit with prompt` icon on an element's toolbar. + + + +You can also attach context to any prompt — a specific data source, a page or element, or an image — so Assistant builds from exactly the right information. Attach a screenshot of a dashboard from any tool, and Assistant rebuilds it natively in Sigma. + + + + + +We take this much further — one-shot dashboard building, editing an existing workbook conversationally, and building from images and plans — in the [Build Dashboards and Apps with Sigma Assistant](https://quickstarts.sigmacomputing.com/guide/aiapps_build_with_assistant/index.html) QuickStart ![Footer](assets/sigma_footer.png) @@ -527,39 +582,13 @@ For more information, see [Build Sigma agents](https://help.sigmacomputing.com/d ![Footer](assets/sigma_footer.png) -## Formula Assistant -Duration: 4 - -The Formula Assistant is available directly in the formula bar of any workbook or data model. It can: - -- **Write a formula** — describe what you want to calculate, and AI suggests the formula -- **Correct a formula** — if Sigma detects an error, AI interprets your intent and suggests a fix -- **Explain a formula** — AI describes what an existing formula does in plain language - -It becomes available when a table is selected: - - - -Notice that Sigma Assistant generated a column called `performance_metric`. It's not immediately obvious what that column is calculating. This is a perfect use case for the Formula Assistant's `Explain this formula` capability. - -Click the `performance_metric` column to select it. In the formula bar, click the AI wand icon and select `Explain this formula`: - - - -The Formula Assistant will return a plain-language explanation of exactly what the formula calculates — what columns it references, how the calculation works, and what the result represents: - - - -While this is a very basic example, this is an important pattern: Sigma Assistant generates analysis quickly, and the Formula Assistant lets you verify what was built. You stay in control of the results, even when AI does the initial work. - -For more information, see [Use AI with formulas](https://help.sigmacomputing.com/docs/use-ai-with-formulas) - -![Footer](assets/sigma_footer.png) - - ## Sigma MCP Server Duration: 5 + + ### What is an MCP server? MCP stands for **Model Context Protocol** — an open standard that allows AI assistants to connect to external tools and data sources in a structured, secure way. Think of it as a universal adapter: instead of an AI assistant being limited to what it already knows, MCP lets it reach out to live systems and work with real data in real time. @@ -568,7 +597,7 @@ The Sigma MCP Server implements this protocol for Sigma. Once configured, Claude ### What this means for Sigma users -The AI features covered in the previous section (Sigma Assistant, Formula Assistant, etc.) work inside Sigma — you open a workbook, ask a question, and the answer appears in context. +The AI features covered so far — Formula Assistant, Sigma Assistant, Building with Assistant, and Sigma agents — work inside Sigma. You open a workbook, ask a question, and the answer appears in context. The MCP Server flips that around. You stay in your AI tool and bring Sigma to it. The core workflow is: @@ -679,7 +708,7 @@ From the Sigma home page, click `Create New` in the left sidebar, then select `W Sigma opens a blank workbook in the exploration state. You’ll notice there’s no name yet — the workbook doesn’t exist in your files until you choose to save it. -