Transformation & the Modern Data Stack

Part 4: Moving & Shaping Data

This group sits in Part 4, Moving & Shaping Data, right after the ingestion and streaming decisions covered earlier in the Part — a deliberate sequence, since once you’ve decided how data lands and how fast, the next question an architect owns is what happens to it after it lands, before anyone downstream can query it with confidence. These two topics work as a single argument: the first settles where transformation compute should run (ETL vs. ELT) and how the resulting layers are organized (the medallion pattern); the second settles how the transforms themselves are written, contractually enforced, and safely rerun without corrupting what’s already landed. Together they turn “the T in ETL/ELT” from a vague step in a diagram into a concrete, version-controlled, testable part of the platform.

mindmap
  root((Transformation & the Modern Data Stack))
    ETL vs ELT & the Medallion Pattern in Practice
      Bronze / Silver / Gold layering
    The dbt Paradigm: Transformation as Code, Data Contracts & Idempotent Reprocessing
      Idempotent reprocessing

See also: Medallion Architecture: Bronze/Silver/Gold for the architectural treatment of the layering pattern this group applies operationally.

Topics


| ← Previous: Should This Be Streaming At All? RT vs NRT Trade-offs & Exactly-Once Semantics | Next: ETL vs ELT & the Medallion Pattern in Practice → | |:—|—:|


Table of contents


Back to top

Independent, self-authored data architecture field notes.

This site uses Just the Docs, a documentation theme for Jekyll.