Platform · Sigma
The export was always step one. Delete it.
Sigma puts a spreadsheet-familiar interface on live Snowflake data. Explore, change inputs, write results back, and trigger the next action without leaving the view or copying the data.
Different ways of working, not a longer feature list.
The familiar grid, pivots and formulas people already know, running directly on the warehouse, at billions of rows, with no download to age and no version to reconcile.
Input tables and write-back mean the decision is recorded where it is made. Scenario, forecast and approval all happen on the platform instead of in an email thread.
Warehouse-native access means the whole business can work with governed data without a second copy to secure or a per-seat gate on curiosity.
Datasets and metrics are defined once, in Sigma, on Snowflake. Reports, apps and agents all read the same definition, so the number matches wherever it appears.
Ask Sigma and Sigma Assistant: an AI analyst, not an AI answer.
Sigma's AI does not hand you a screenshot and a shrug. It works the way an analyst works: exploring, drilling and building, and leaves you a live, governed workbook you can keep using.
Ask Sigma
An AI analyst that works like an analyst
Ask in natural language and Ask Sigma deploys the agent trained for that question, explores the live warehouse data, drills into the detail and builds the workbook, editable and traceable to the source.
- Natural-language queries over live Snowflake data
- Returns an editable workbook, not a dead answer
- Every result traceable to the warehouse
- Discover mode for open-ended exploration
Sigma Assistant
One governed AI surface
for analysing and building
Introduced in 2026, Sigma Assistant unifies Ask Sigma and AI Builder: analyse data and build applications or workbooks in natural language, all against your real business data and permissions.
- Build apps and workbooks from a prompt
- Analysis and creation in one place
- Bound to your data and your governance
- Custom agents via the Chat element
Cortex Agents in Sigma
Snowflake's agents, called into the workbook
Sigma can invoke a pre-built Snowflake Cortex Agent directly, enrich the data, then extend the agent's output into a working AI application. One governed model, two surfaces.
- Warehouse-native agents inside Sigma
- Enrich data without leaving the workbook
- Output extends into an app
- Governed once, at the warehouse
Not the feature list. The reasons Sigma on Snowflake pays back.
Transformation
The export stops being the first step of every analysis. Work that used to leave the platform now happens on it, on governed data, in a form the next person can pick up.
Business process improvement
Analysis and action share a surface. Write-back and input tables turn a dashboard into a place where planning, approval and correction actually get done.
Optimisation
One governed model replaces the same logic rebuilt in dozens of reports. Fewer definitions to maintain, fewer numbers that disagree, less analyst time spent reconciling.
Cost saving
Warehouse-native and usage-based: no second copy of the data to store and secure, and no per-seat rationing that keeps people off the platform.
Future-proofing
The same platform grows from dashboard to input-driven data app to embedded analytics to agents, without changing tools or moving the data.
Sigma, delivered on a governed Snowflake foundation.
Sigma is only as trustworthy as the model underneath it. We implement it where it belongs: on a warehouse that is already modelled, secured and trusted.
Certified delivery partner
Certified by Sigma as a delivery partner, recognised for the technical depth to implement it properly rather than just switch on a licence.
Sigma partner and reseller
Licences, implementation and account management run through one partner, with the semantic layer Sigma reads from built on Snowflake by the same team.
Adoption, not just access
Adoption programmes that move the spreadsheet habit onto live data, so the licence turns into a changed way of working, not a login nobody uses.
South Africa's home team
Local expertise, local context, local accountability. Platform and ways of working delivered together, for the long term.
