← Back to home

Truly generic

Classifications, personas, templates, and outputs are all yours to define. ADL ships starter configurations as a head start — the framework underneath stays yours to shape. Methodology, taxonomy, perspective, output: nothing is hardcoded.

Why this matters

The deepest source of vendor lock-in isn't the database your metadata sits in; it's the assumptions the tool makes for you. Tools that hardcode a methodology (only Data Vault, only Kimball) leave you stranded when the business needs something else. Tools that hardcode a taxonomy (PII is the only sensitivity tier; Hub/Link/Satellite are the only object types) leave you fighting the tool when your governance needs a richer model. Tools that hardcode a single audience view leave engineers and compliance reviewers seeing the same crowded screen, neither of them well served.

Every one of those decisions is a future tax: each time the business evolves, the tool needs to be coaxed, worked around, or replaced. Truly generic means the tool doesn't ship those decisions in the first place. It ships the framework underneath, and you supply the rules that fit your work.

How ADL delivers

ADL is configurable on four independent axes. Each is something your team defines, not something ADL prescribes.

Structure

The metadata model itself — Data Objects, Data Items, Connections, Mappings, Relationships. Captures any kind of design, from business capabilities to physical schemas.

Taxonomy

Classifications and Conventions you define. PII / Sensitive / Confidential for compliance; Hub / Link / Satellite for Data Vault; Strategic / Tactical / Operational for capability modeling — your call.

Perspective

Personas and visualization rules filter the graph for the audience reading it. The compliance reviewer sees PII chips; the engineer sees layer markers; the architect sees the conceptual hierarchy. Same metadata, different lenses.

Activation

Templates produce any text-based output. SQL, documentation, deployment scripts, test fixtures, API configurations, custom report formats. If it's text-based, ADL can generate it.

The starter configurations are reference, not constraint

ADL ships sensible defaults across all four axes — the starter classifications cover Conceptual / Logical / Physical groupings; the template library covers Data Vault, Persistent Staging, and documentation; the sample designs show how the four axes are configured together. None of them are the limit.

If you can describe how your design should be structured, classified, viewed, and rendered, ADL can carry it. The framework is what we ship; the configuration is yours.

What ships today

  • Open framework, configurable everywhere. Classifications, conventions, personas, templates — all editable, all replaceable, all version-controlled in your repo.
  • Reference configurations across all four axes. 8 starter sample designs (Data Vault VDW/PDW, PSA across Snowflake/SQL Server/Fabric, plus NBility, Metadata Architecture, BusinessModel); 20+ production templates; 13 Handlebars helpers; a global classifications reference set.
  • The philosophyAgnostic philosophy on the blog explains the why behind the design.

Try the app — free in public preview