Phase 01 · Foundations · sql

SQL from Zero to Advanced

How you ask a warehouse questions, and how you shape it so the answers are right.Start with what data and databases are, then learn SELECT, JOINs, GROUP BY, window functions and star schemas, with a fact-table capstone, all in the SQL editor.
Start lesson 1What is data?
lessons
33
modules
6
stages
5
of learning
6h 29m

Concept Traces in this track

Why this track exists

Imagine the business asks a simple question: what was revenue last month, by country, excluding refunds. The data for that answer sits in 4 tables written by three different teams. One of those tables has 2 rows per order because a retry inserted a duplicate. If you get the join wrong, you report a number that is twice as large as reality, and somebody makes a decision on it. SQL is the language for getting that right, and warehouse modelling is how you stop the same mistake recurring every month.

SQL is the one skill that appears in every data engineering interview and every working day. You will read more SQL than you write, most of it written by somebody who has left the company, and you will be asked why a dashboard number changed.

What you need before starting

  • No database experience

    The track opens with what data and a table actually are, before any query appears.

  • Helpful but optional: basic Python

    Not required for any lesson here. Queries run in the SQL editor in your browser.

The roadmap

5 stages, in the order they build on each other. Each stage lists the modules and lessons it covers, and what you should be able to do by the end of it.

01What data and tables are

Terminology like row, column, primary key, and grain gets used casually by everyone around you. If those words are fuzzy, every later lesson is guesswork.

Understanding DataFree

0/4

What data is, what databases are, and why tables have rows and columns. The vocabulary you need before writing SQL.

  1. What is data?8m
  2. What is a database?8m
  3. Tables, rows, and columns10m
  4. Primary keys and foreign keys10m

Module checkpoint

1 conceptual questions to verify mastery

By the end of this stage

You can look at a table and say what 1 row represents, which is the single most useful habit in this whole track.

02The relational core

90 percent of working SQL is SELECT, WHERE, GROUP BY, and JOIN. The goal here is not exposure, it is fluency, because these are the tools you use every day.

Relational coreFree

0/6

SELECT, WHERE, GROUP BY, and JOIN: the four operations every SQL query uses.

  1. Relational basics8m
  2. NULL handling10m
  3. Aggregations & grouping10m
  4. Multi-table joins12m
  5. Subqueries & CTEs10m
  6. SQL that changes data: CREATE, INSERT, UPDATE, DELETE12m

Module checkpoint

3 conceptual questions to verify mastery

By the end of this stage

You can answer a business question that spans 3 tables and defend why your join does not duplicate rows.

Where people get stuck

Joins are where beginners produce confidently wrong numbers. If a join makes your row count go up unexpectedly, stop and find out why before continuing.

03Windows, cleaning, and harder questions

Some questions cannot be answered by collapsing rows: rank customers within each region, find the running total, compare each row to the previous one. GROUP BY throws away exactly the rows you need.

Windows, cleaning & opsFree preview

0/5

The DE benchmark: window functions, cleanup, dedupe, and reading EXPLAIN.

  1. Window functions (the DE benchmark)Free14m
  2. Data cleaning & string/date manipulation10m
  3. CASE expressionsFree10m
  4. Deduplication & incremental upserts12m
  5. Indexing, partitioning & optimization12m

Module checkpoint

1 conceptual questions to verify mastery

By the end of this stage

You can write window functions without looking up the syntax, and you know when a CTE makes a query readable rather than just longer.

Where people get stuck

NULL behaviour breaks more queries than complex logic does. NULL is not zero and not empty, and it makes both comparisons and aggregates behave in ways that surprise people.

04Modelling a warehouse

Querying well is not enough. If the tables themselves are shaped badly, every analyst will keep double-counting revenue and you will keep explaining why. This stage is how you shape them.

By the end of this stage

You can design a fact table and its dimensions, state the grain out loud, and handle a dimension whose values change over time.

Where people get stuck

Slowly changing dimensions are the classic interview topic here, and the classic real-world mess. Learn type 1 versus type 2 properly.

05Performance and production

A query that works on 10,000 rows and dies on 10 million is not finished. This is where you learn to read what the engine is actually doing.

By the end of this stage

You can read a query plan, explain why a query is slow, and say which change will help before you make it.

Finish this track, get a certificate

Complete all 33 lessons and claim a certificate with a link you can add to LinkedIn or send to a recruiter.

How certificates work

Mistakes to avoid on this track

Common mistakes on this track and what to do instead
Common mistakeWhat to do instead
Memorising syntax without ever checking row counts.After every join, check whether the row count changed as you expected. This one habit prevents most reporting errors.
Treating SELECT DISTINCT as a fix for duplicates.DISTINCT hides a join problem rather than solving it. Find where the extra rows came from.
Learning window functions as three syntax examples.Learn the mental model: the window stays open, the rows survive. Then the syntax follows on its own.
Skipping the modelling module because querying feels more practical.Modelling is what senior interviews probe. Anyone can write a SELECT, few can defend a grain.

Where to practise this

Where to go after this

SQL from Zero to Advanced reviews & rating

Finished a few lessons? Tell us what worked and what was confusing. Reviews are read by hand and help decide what gets rewritten next.