# Sigma

> A spreadsheet over the warehouse, with write-back

Sigma is a cloud analytics platform whose interface is a spreadsheet: business users pivot, filter, and add formulas against live warehouse tables, with input tables writing values back into the warehouse rather than into a local file.

Website: https://www.sigmacomputing.com

## Use it when

- Business users live in spreadsheets: formulas, pivots, and filters run directly against live warehouse tables, with no extract and no local copy.
- Write-back is the point: input tables put forecasts, adjustments, and approvals back into warehouse tables, turning reporting into a lightweight application.
- A cloud warehouse is already in place: Snowflake, Databricks, BigQuery, or Redshift is the platform Sigma presumes.
- Many consumers, few builders: consumption licences cost a fraction of a Build seat.

## Think twice when

- There is no cloud warehouse: no file uploads and no SaaS APIs, so without one it is not a candidate.
- Public pricing matters: licensing is quoted through sales.
- Row-level security must be airtight: the docs warn the filter formula on a table can be modified by viewers, so the protection really covers downstream elements.

## How it runs

Managed SaaS that pushes every query down to the connected platform; materialization writes copies back into it. Embedding uses JWT-signed URLs minted by your backend, carrying user attributes so data-model policies apply; version tags drive git-based CI/CD through the REST API.

## Details

### General

| Attribute | Value |
| --- | --- |
| License | Commercial |
| Deployment | Managed |
| Workload | Interactive |
| Operational complexity | Low |
| Pricing | Subscription |
| Documentation | https://help.sigmacomputing.com/docs |

### Business Intelligence

| Attribute | Value |
| --- | --- |
| Authoring model | A spreadsheet (formulas, pivots and filters) or SQL |
| Semantic layer | Data models: reusable elements, relationships, metrics and access controls |
| Connectors | Cloud data platforms only: Snowflake, Databricks, BigQuery, Redshift, Starburst, ClickHouse, Postgres, MySQL, AlloyDB, Azure SQL and SQL Server; no file uploads, no SaaS APIs |
| Query execution | Pushdown to the connected platform; materialization writes a copy back into it |
| Chart library | Built-in chart types, extended by custom plugins built against the Plugin Development API and registered per organisation rather than pulled from a public marketplace |
| Definitions in | Sigma's cloud; version tags drive Git-based CI/CD through the REST API |
| Access control | Row-level security as a formula column on the data model, keyed on user attributes, team membership or email, plus column-level security, but the docs warn the filter can be modified by anyone viewing the table, so the protection really covers child elements and downstream sources |
| Alerts and delivery | Conditional exports (a scheduled export that fires only when the condition holds) to email, Slack, Teams, SharePoint, Google Sheets or Drive, a webhook or cloud storage |
| Embedding | Public embeds for workbooks, pages and elements, or secure embeds where a server-side API mints a one-time JWT-signed URL and the host application's own user security applies; data models and reports require the secure kind |
| Billing unit | Per user by licence type (View, Act, Analyze and Build) with nothing published; Sigma quotes through sales, and the cheap end is where its argument sits, since consumption licences cost a fraction of a Build seat |

## Capabilities

- [Business Intelligence](https://matca.io/capabilities/business-intelligence)

---

Source: https://matca.io/sigma
Last updated: 2026-08-31T11:40:49.692Z