Transformation Framework
also called ELT Framework, SQL Transformation Tool
A tool that turns warehouse transformations into version-controlled, tested, dependency-aware code rather than a collection of scheduled scripts.
The shift these tools represent is applying ordinary software engineering to the transformation layer — which for two decades was the least engineered part of most data estates: SQL in a scheduler, no tests, no version control, no lineage, and no way to reproduce last quarter's numbers.
What the framework provides is modest individually and transformative together: transformations as files in a repository with review and history; a dependency graph inferred from references rather than maintained by hand; environment separation so a developer builds into their own schema against the same code; data tests as assertions that run with the build; and documentation and lineage generated from the code rather than written separately and left to rot.
The pattern it encourages is ELT: load raw, transform in the warehouse using its compute. That is usually right on modern platforms and is not universal — very large transformations and non-SQL work still belong in a processing engine.
The failure to watch for is model sprawl. Without deliberate layering and ownership, teams accumulate thousands of models with deep dependency chains where a change anywhere requires rebuilding half the graph, and the runtime and cost grow past what anyone budgeted.