Learn Satsuma

Choose a path based on your role. Whether you review mappings, build pipelines, or automate with AI, there is a learning track for you.

Quick Start

Up and running in 4 steps

From zero to your first validated mapping in under two minutes.

1

Install the CLI

# Universal — works on macOS, Linux, and Windows
npm install -g https://github.com/thorbenlouw/satsuma-lang/releases/download/v0.7.0/satsuma-cli-v0.7.0.tgz

Requires Node.js 18+. One universal package works on every platform (WASM-based, no native compilation). Available from the v0.7.0 release. Also available: latest unstable build.

2

Install the VS Code extension

# Download from the latest release, then:
code --install-extension vscode-satsuma-v0.7.0.vsix

Download vscode-satsuma-v0.7.0.vsix from the v0.7.0 release. Also available: latest unstable build. Get syntax highlighting, real-time diagnostics, go-to-definition, IntelliSense, lineage visualisation, and more.

See all extension features →
3

Create your first .stm file

schema source {
  name  STRING
}

schema target {
  full_name  STRING
}

mapping {
  source { `source` }
  target { `target` }

  name -> full_name
}

Define schemas, then connect them with a mapping block. The -> arrow reads as "maps to".

4

Validate with the CLI

satsuma validate my-mapping.stm

The parser checks syntax, reference resolution, and structural correctness. Zero errors means you are ready to go. The VS Code extension also shows diagnostics in real time as you type.

Explore all 21 commands →

Choose your learning path

Every role interacts with data mappings differently. Find the resources that match your workflow.

Product Owners & Analysts

Product managers, product owners, and business analysts — read, review, and sign off on data mappings without learning to code.

Key Concepts

  • Schemas and field definitions
  • Mapping blocks and arrows
  • Natural language descriptions
  • Metadata and annotations

What You'll Learn

  • Reading and reviewing mappings
  • Reviewing changes in pull requests
  • Understanding data lineage
  • Validating business rules

Data & ML Engineers

Build, validate, and trace data pipelines with parser-backed tooling.

Key Workflows

  • validate Check syntax and references
  • lint Enforce conventions with auto-fix
  • lineage Trace field-level data flow
  • graph Visualize workspace structure
  • diff Compare mapping versions

Example Walkthroughs

The examples gallery includes real-world patterns: multi-source joins, SCD Type 2, Kimball star schemas, and more.

Browse examples →

Integration Engineers

Map complex interchange formats: EDI, XML, COBOL copybooks, Protobuf, and more.

Format Focus

  • EDI X12 / HIPAA
  • COBOL Copybook
  • HL7 / DICOM
  • ISO 20022 / SWIFT MT
  • FIX Protocol / Protobuf

Why Satsuma?

Satsuma handles deeply nested structures, repeating segments, and format-specific idioms natively. Each schema format convention documents how to represent the format's structures in Satsuma.

Governance & Audit

Trace every field from source to target. PII tags, encryption markers, and lineage are structural.

Focus Areas

  • PII tags and sensitivity markers
  • End-to-end lineage tracing
  • Ownership, retention, masking
  • Regulatory compliance (GDPR, HIPAA)

Key Commands

  • lineage Trace data flow end-to-end
  • where-used Find all consumers of a field
  • find --tag Search by metadata tag
  • graph Dependency visualisation

Example

The governance example demonstrates owner, steward, classification, retention, masking, encryption, and compliance tokens. See also the Governance Convention Guide.

View governance examples →

AI & Automation Engineers

Build reliable AI workflows with a constrained grammar that LLMs can generate and validate.

Primary Resource

satsuma agent-reference

Also available via the CLI for piping into agent prompts.

Key Features

  • Compact EBNF grammar
  • Deterministic parsing via tree-sitter
  • Validation-in-the-loop workflows
  • LLMs generate valid Satsuma >90%

Hybrid Workflows

Use AI to generate initial mappings from source documentation, then validate with the CLI, review with humans, and iterate. The constrained grammar means fewer hallucinations and reliable round-tripping.

Enterprise Architects

Model entire data platforms with namespaces, imports, and workspace-wide analysis.

Focus Areas

  • Namespace-scoped platform models
  • Cross-file imports and references
  • Workspace-wide dependency graphs
  • Multi-team mapping ownership

Platform Modelling

Use a platform entry point file with import statements to pull definitions across domains. Then use satsuma graph to visualize the whole platform.

Documentation Hub

All documentation in one place

Specifications, references, tooling guides, and technical deep-dives.

Satsuma

Ready to start mapping?

Satsuma is open source and free. Install the CLI and write your first mapping in minutes.