Learn the Pattern, Not the Product
Data engineering fundamentals that transfer across Snowflake, Databricks, BigQuery, and Microsoft Fabric — because the durable skill is the pattern, not the platform.
Why Patterns Over Products
The data industry is converging on unified, PaaS-style platforms — Microsoft Fabric, Databricks, Snowflake, and Google BigQuery. Each one markets its own menus, brand names, and "magic" buttons. But underneath, they all rearrange the same handful of patterns: ingest data, store it cheaply, compute on it elastically, model it, serve it, and govern it.
Memorizing one vendor's interface is a depreciating asset — the UI changes, and your knowledge expires with it. Understanding the underlying pattern is a transferable one: learn it once, and any platform becomes navigable. That is the entire premise of this series. Every concept is taught once, then mapped — in a "Same pattern, every platform" table — across the four leading platforms, so the knowledge moves with you.
The rule of thumb: when you meet a new data platform, don't ask "where are the buttons?" Ask "which of the six layers is this, and what pattern is it implementing?" The answer is almost always one you already know.
The Six Layers Every Platform Shares
Whatever the logo on the login screen, a data platform is these six layers stacked on top of each other:
Storage and compute are separated (Layers 3 & 4); governance (Layer 6) wraps the whole stack. Part 1 of the series unpacks this map.
What This Series Is — and Isn't
Pattern-first
Each post teaches one concept, then shows how it appears across four platforms. You learn the idea, not the menu.
Explained in 60 seconds
Every concept opens with a plain-language, numbered breakdown — no slides, no jargon wall. Then a deeper dive for those who want it.
Vendor-neutral
No platform is "the answer." The goal is knowledge you can carry into Fabric, Databricks, Snowflake, or BigQuery without starting over.
The 12 Concepts
Two posts a week, six weeks. Each is self-contained, but together they form a complete mental model of a modern data platform:
Anatomy of a Data Platform
The six universal layers behind every platform.
Storage vs Compute
Why every modern platform separates them.
ETL vs ELT
Where transformation actually lives.
Batch vs Streaming
Choosing your latency on purpose.
Data Modeling 101
Normalized for writes, analytical for reads.
Dimensional Modeling & the Star Schema
The 30-year-old idea still powering BI.
Lake vs Warehouse vs Lakehouse
The convergence shaping the industry.
Open Table Formats
Delta, Iceberg, Hudi — and the end of lock-in.
Partitioning & Clustering
The universal lever for speed and cost.
Data Quality & Testing
Trust before dashboards.
Orchestration & CDC
Keeping data fresh, hands-off.
Governance, Catalog & Lineage
Who, what, and where — the trust layer.