Skip to main content
Sigma Computing

It’s time to migrate from Hex to Sigma

Analysis that only the people who wrote the notebook can change? Sigma rebuilds your Hex projects as live, governed workbooks where business users explore in a spreadsheet and analysts keep SQL and Python.

Schedule a demoStart a free trial
claude · sigma-migration-skills

Teams of every size are moving from Hex to Sigma

DoorDashDuolingoAB CarValAffirmVantaConagraWorkday

How the migration works

Three steps from Hex to Sigma

Point Sigma at your warehouse and it rebuilds your projects as live, governed analytics your whole team can use.

  1. Connect your warehouseSigma queries Snowflake, Databricks, BigQuery, or Redshift live, so nothing gets extracted and no data leaves your warehouse.
  2. Convert your Hex projectsThe Hex converter rebuilds SQL cells, metrics, charts, and app layout as a Sigma data model and workbook, then checks the numbers against your warehouse.
  3. Explore and extendAnyone who knows a spreadsheet can explore, write back, and build AI Apps on top, with row- and column-level security inherited from your warehouse.

Sigma Migration Skills

The Hex migration runs in your coding agent

An assessment inventories and scores your Hex content first. The converter rebuilds it in Sigma and checks the results against your warehouse. It’s open source and works with Claude Code, Cursor, Snowflake CoCo and other coding agents.

/plugin marketplace add sigmacomputing/sigma-migration-skills
/plugin install sigma-authoring@sigma-migration-skills
/plugin install hex-to-sigma@sigma-migration-skills
Migrate this Hex project to Sigma
View the skill on GitHub Follow the QuickStart

Where your data goes to work

Most BI tools stop at the dashboard. Sigma keeps going, turning the same governed data into interactive apps and agents that act.

  • Semantic portability

    Protect existing investments. Sigma natively integrates with dbt, Snowflake Semantic Views, and Databricks Unity Catalog metric views, so externally built metrics flow in without re-definition.

  • Sigma Agents

    Turn insights into automated work. Sigma Agents read, write, and trigger external workflows while inheriting warehouse security, so every action stays auditable.

  • AI Apps

    Move beyond read-only dashboards. Anyone can build interactive, no-code AI Apps that take action and write decisions directly back to your warehouse.

  • Secure governance

    AI security must be architectural. Sigma Agents and AI Apps automatically inherit your cloud data warehouse row-level and column-level security.

  • Enterprise SDLC

    Get production-grade controls without the engineering overhead. Sigma isolates draft and live states using connection-aware deployment and version tagging.

  • AI ecosystem

    Sigma is your OS for live data and AI. Unify external agents via MCP with warehouse LLMs and Sigma Agents for natural-language discovery, without vendor lock-in.

Sigma is the enterprise leader in self-service analytics and operational workflows

As of March 30, 2026.

See the full Sigma vs Hex comparison
Architectural ComplexityHexNo. Relies on a stateful, kernel-driven notebook architecture with linear, fragile execution. Complex analysis requires loading data into memory-constrained dataframes, forcing manual management of kernel states, execution order, and compute bottlenecks.SigmaYes. One architecture: Live query against the cloud data warehouse. Everything from ad-hoc analysis to Sigma Assistant and Sigma Agents run natively on your existing, governed semantic models without duplicating cloud infrastructure.
Required skillsHexNo. Demands technical fluency. The underlying mechanics require a strong understanding of dataframes, SQL, and Python. Users that do not know how to code must rely on data scientists or developers to build or modify any dashboard.SigmaYes. Sigma empowers domain experts to build Sigma Agents and AI Apps using familiar functions in a spreadsheet UI alongside standard SQL and Python. Teams deploy functional agents and closed-loop apps in days without data engineering bottlenecks.
Multi-Modal DevelopmentHexNo. Forces a linear, cell-based notebook workflow. Architected for data science narratives, this rigid, top-to-bottom structure intimidates non-technical users and makes multidimensional, ad-hoc exploration incredibly cumbersome.SigmaYes. Unifies technical and non-technical users by enabling them to co-create across spreadsheets, SQL, Python, and AI in one workspace. No context-switching, no duplicate data, and no broken governance.
Collaborative WorkflowsHexNo. Multiplayer collaboration exists, but is practically restricted to technical users who understand notebooks. Business users are relegated to being passive consumers of dashboards, breaking the synchronous co-creation loop between technical and domain experts.SigmaYes. Delivers synchronous, multi-user state management and cross-functional collaboration where business users and data teams co-edit live workbooks together. This shared environment natively supports agile approval chains and real-time operational execution.
SecurityHexNo. Breaks native warehouse security perimeter because data is extracted into Hex’s compute kernels for processing. Admins need to manually rebuild and maintain duplicate security models within Hex since native governance policies (RLS/CLS) do not automatically apply. SigmaYes. Sigma's zero-copy architecture natively inherits warehouse Row-Level Security (RLS) and Column-Level Security (CLS) via OAuth. No extraction or duplicate permission models to manage. Every user, Sigma Agent, and AI App operates safely under your centralized governance model.
Enterprise ScalabilityHexNo. Struggles to scale due to dataframe memory limits, complex compute profiles, and a steep learning curve that inhibits adoption among anyone unfamiliar with code.SigmaYes. Easily scales to tens of thousands of concurrent business users conducting massive, ad-hoc explorations directly against the cloud data warehouse without compute bottlenecks.
AI Model FlexibilityHexNo. Lacks a no-code LLM routing layer. To change models, a data scientist must manually rewrite the API integrations within the Python backend, stalling agility and driving up costs. Business users are locked into technical administrator configurations, unable to switch models within workflows to optimize for cost and performance.SigmaYes. LLM-agnostic architecture lets you directly connect to OpenAI, Azure OpenAI, or Google Gemini and use any Snowflake Cortex or Databricks hosted LLM including Claude from Anthropic. Future-proof your infrastructure and easily switch models using familiar spreadsheet functions to optimize for cost and performance without needing code to rebuild your workflows.
Primary User ParadigmHexNo. A cloud-native notebook for data teams and coders that excels at Python modeling but relies on a linear, cell-based paradigm. It fundamentally alienates business users, keeping them dependent on data teams for answers.SigmaYes. Built for enterprise. Sigma's warehouse-native spreadsheet UI scales to billions of rows, empowering all users to conduct ad-hoc analysis, build AI workflows and agents, and execute decisions without code.

Straight from the source

Sigma on G2 (opens in a new tab)Sigma on Gartner Peer Insights (opens in a new tab)
  • “Created for those with SQL, Python, & Excel knowledge. Allows for transferring those skills directly over to the platform without learning some domain-specific modelling language (LookML, DAX, M, etc.)”

    Finance AssociateSoftware · Gartner Peer Insights (opens in a new tab)

  • “I like many of the features in Sigma, but my top three features would be the ability to write in SQL or Python, write back functionality with a spreadsheet like interface, and the live chat feature. The ability for technical users to write custom SQL or Python and allow end-users to further use those results is a game changer.”

    Data and Analytics ManagerManufacturing · Gartner Peer Insights (opens in a new tab)

  • “Sigma has been very easy to set up and use for both advanced and less technical data users. Power users can work directly with SQL, while newer users can take advantage of the built-in help and AI features to quickly understand the data and get productive.”

    Simon IdrisVP Engineering · Mid-Market (51-1000 emp.) · G2 (opens in a new tab)

Schedule a live 30-minute demo

See your own Hex projects rebuilt as live, governed Sigma workbooks in a 30-minute walkthrough on your data.

Schedule a demoStart a free trial