DataMuru

Provider-agnostic data infrastructure, governed by design.

DataMuru is an Apache-2.0, Python-first data infrastructure framework. It provisions and governs provider-backed data estates from declarative configuration. This public repository is the canonical home of the DataMuru Open Source Edition, the shared configuration contract, the datamuru PyPI package, and the public documentation.

This repository contains the v0.6 alpha implementation with the following scope:

• Foundation layer: config loading, validation, local state, planning, apply, and destroy.
• Azure-first Databricks provider abstraction with multi-cloud-ready interfaces.
• Basic governance compilation: taxonomy, RBAC, and masking integration points.
• Core CLI surface: init, validate, plan, apply, and destroy.
• MkDocs-based product documentation written from an international SaaS product perspective.
• Live Databricks catalog, schema, ACL, import-discovery, and supported identity workflows.
• Provider-neutral Databricks/Snowflake metadata inventory evidence with stable IDs.

Current stage

DataMuru is currently in the v0.6 alpha stage.

The package and CLI execute supported live Databricks operations when execution_mode: live-apply is configured. Alpha support currently covers catalogs, schemas, Unity Catalog ACLs, import discovery/config generation, and capability-aware account SCIM identity operations.

Delivery status

Product execution is tracked in a private DataMuru Product Roadmap GitHub Project. The repository snapshot is maintained in PROJECT_STATUS.md, including readiness, risks, milestones, completed work, blocked work, and next recommended actions.

Open-core model

• DataMuru OSS: this public Apache-2.0 repository and PyPI package.
• DataMuru Enterprise: a private extension repository for paid capabilities such as multi-workspace orchestration, advanced compliance automation, hosted services, SSO/SAML, SIEM integrations, and SLA-backed operations.

Enterprise extends the public core; it does not maintain a competing fork of the core configuration model or CLI.

Design constraints

• Multi-cloud is an architectural requirement, but not a parity requirement for the alpha slice.
• open-source vs enterprise is the only runtime packaging boundary.
• Zero external data-tool runtime dependencies means no Terraform, dbt, Airflow, Great Expectations, Fivetran, or Airbyte required for the framework to operate.

Quick start

pip install "datamuru==0.6.0a0"
datamuru init --name quickstart --output-dir datamuru-quickstart
cd datamuru-quickstart
datamuru validate --config datamuru.yml
datamuru doctor --config datamuru.yml
datamuru inventory export --config datamuru.yml --out .datamuru/metadata-inventory.json --output json
datamuru plan --config datamuru.yml

Installation

From PyPI:

pip install datamuru

Documentation

This repository now includes a full MkDocs documentation site.

• MkDocs config: mkdocs.yml
• Docs source: docs
• Published documentation: ajayaj2000.github.io/DataMuru
• Support and documentation feedback: Support and feedback

To work on the documentation locally:

python -m pip install -e ".[docs]"
python -m mkdocs serve

Product usage guidance

For operator guidance and rollout practices, start with:

• Product Usage Guidelines
• Team Adoption Guidelines

Trying it with Databricks Free Edition

If you want to try the framework with your own Databricks personal workspace, start here:

• Databricks Free Edition Setup

For package-oriented team usage, also read:

• Packaging and Distribution
• Product Usage Guidelines

Repository standards

• datamuru/: shared installable Python package
• docs/: versioned product documentation published through GitHub Pages
• examples/: shareable sample DataMuru projects and generated-project fixtures
• schemas/: public configuration contracts
• tests/: unit, provider-contract, and end-to-end tests
• .github/workflows/: CI, documentation deployment, link validation, and trusted PyPI publishing

See CONTRIBUTING.md for the required local quality gate.