SQL for Every Data Platform
Before Fabric, Snowflake, Databricks, or BigQuery, there's one skill every one of them assumes: SQL. This series starts at the fundamentals and builds to procedural SQL and cross-platform syntax — every concept explained, runnable, and mapped across platforms.
Why SQL Is Still the First Skill
Every platform in this series — Snowflake, Databricks, BigQuery, Fabric, Redshift — speaks SQL as its primary interface. The dashboards, the dbt models, the ad-hoc analysis: all SQL underneath. Learn the language once, and the platform becomes a detail — an analyst who is fluent in the concepts below can sit down in front of any of these five platforms and be productive within the hour, translating unfamiliar keywords instead of re-learning how relational data works.
This series is built as a working reference, not a quick skim: every concept — from COUNT/SUM/AVG to Slowly Changing Dimensions to stored procedures — is explained in full, with a runnable example, before moving to the next. All 14 companion notebooks execute end to end with zero errors, and every "same query, every platform" claim is flagged for its confidence level, not presented as fact to memorize.
The rule of thumb: learn a concept once here, in SQLite/DuckDB, with real executed output — then Part 12 and the per-topic notes throughout the series show you where the same idea lives on BigQuery, Snowflake, Databricks, Fabric, and Redshift, so the knowledge transfers instead of expiring with one vendor's syntax.
From Fundamentals to Platform Fluency
The 14 parts build in four stages, each one assuming the last:
Structure and syntax first, then day-to-day querying, then production-grade patterns, then cross-platform fluency.
What This Series Is — and Isn't
Reference-depth, not a skim
Every concept is explained in full — what it is, why it matters, the common pitfall — with a runnable example, not a one-line definition.
Runnable, verified code
All 14 companion notebooks execute end to end with zero errors, including a real dbt project and measured (not claimed) performance results.
Platform-agnostic
Learn each concept once in SQLite/DuckDB, then see exactly where it lives on BigQuery, Snowflake, Databricks, Fabric, and Redshift.
The 14 Parts
Each part is self-contained and independently useful, but together they form a complete reference — from your first CREATE TABLE to comparing five cloud warehouses side by side:
Fundamentals: DDL, DML, DQL
The three pillars: define structure, write data, read it back.
Designing the Schema: Star vs. Snowflake
Fact and dimension tables, the analytical shape of your data.
Joins & Aggregations
COUNT, SUM, AVG, MIN, MAX, and the JOIN types that combine tables.
Subqueries, CTEs & Window Functions
Ask harder questions without losing row-level detail.
Data Cleaning & Formatting with SQL
Dates, rounding, labels, and NULLs — the unglamorous work.
Semi-Structured Data & JSON in SQL
Query nested, schema-less data directly with SQL.
The Query Cookbook
Eleven day-to-day patterns, copy-paste ready.
Slowly Changing Dimensions & Incremental Loading
The interview question about tracking history over time.
Query Performance & Tuning
Indexes, partitioning, views, and a real measured pruning demo.
Simulating a Platform Locally
Run dbt, tests, and governance on your laptop, no warehouse account.
Procedural SQL
Stored procedures, triggers, and PL/pgSQL, PL/SQL, T-SQL compared.
Same Query, Every Platform
BigQuery, Snowflake, Databricks, Fabric, and Redshift syntax side by side.
Choosing Your Database
OLTP vs. warehouse vs. embedded, with pros, cons, and real scenarios.
SQL Style & Best Practices
Naming, safe DML, and a code-review checklist that closes the series.
Runnable code for the whole series: sql-for-data-platforms on GitHub — 14 independently runnable Jupyter notebooks, a working dbt-duckdb project, and an optional Docker Postgres setup for Part 11.
← Back to Publications