Azure Synapse Analytics Course: Free Step-by-Step Project

Learn Azure Synapse Step By Step With Project is a free video course that teaches Azure Synapse Analytics end to end by building a data project. It is for data engineers, SQL developers and BI professionals who want to query a data lake with SQL, build pipelines and Spark notebooks, and load a data warehouse in Azure.

The course has 120 videos and about 11.6 hours of content on YouTube. Lessons are short and practical, usually 2 to 11 minutes, with learning tasks and assignments along the way. You start with the Synapse workspace and Studio hubs, work through a project on a medallion architecture, and finish with Spark, Synapse Link and the dedicated SQL pool. See the full list on the Azure Synapse course page.

Who this course is for

• Data engineers who need to work with data lakes and warehouses in Azure.
• SQL developers who want to use their T-SQL skills on files in a data lake.
• BI developers who report on large datasets with Power BI.
• Learners moving from Azure Data Factory who want to see how Synapse brings pipelines, SQL and Spark together.

The level is Intermediate. You should be comfortable with basic SQL. The course explains data warehouse and data lake ideas from the start.

What you will learn

The 120 videos group into seven themes.

1. Synapse basics and the workspace. “Data Warehouse Vs. Data Lakes”, “Inception of Azure Synapse Analytics”, “Create Synapse Analytics Workspace Lab”, and overviews of Synapse Studio and its Data, Develop, Integrate, Monitor and Manage hubs. Knowing the hubs makes the rest of the course easy to navigate.

2. The project. The project dataset and file types, importing the project data to the data lake, the project requirements and “Synapse Project - Solution Architecture Overview (Medallion Architecture)”. Every later section adds a piece of this solution.

3. Serverless SQL pool. The largest theme: serverless SQL pool introduction and cost control, the OPENROWSET function, data types and collation, querying CSV, tab-separated, JSON (single line, standard, multi-line and arrays with OPENJSON), Parquet and Delta files, folders and file metadata functions. It then uses these skills for data discovery, duplicate detection, data quality checks, joins and transformations.

4. Data virtualisation. External tables for CSV, Parquet and Delta, handling invalid records, views (including views for JSON files) and partition pruning. This is how you give users clean SQL objects on top of raw files.

5. Transformation. Transforming CSV and JSON data to Parquet, partitioned Parquet output, stored procedures, select statements and views for transformation, with learning tasks and assignments to apply them to the project.

6. Synapse pipelines. Pipeline components, linked services and datasets, delete, script and stored procedure activities, parameters and variables, dynamic pipelines, pipelines for partitioned files, pipeline dependencies, and manual and automated triggers.

7. Spark, Power BI, Synapse Link and dedicated SQL pool. Creating a Spark pool, notebooks, integrating Spark pool with serverless SQL pool and running notebooks in pipelines; connecting Power BI Desktop and building two reports; Synapse Link with Cosmos DB, queried from serverless SQL and Spark; and the dedicated SQL pool, loaded with PolyBase and the COPY command and connected from Azure Data Studio and Power BI.

How the course is organised

• 120 videos, about 11.6 hours in total, opening with a 4-minute course introduction.
• Sections run in this order: introduction and workspace, project setup, serverless SQL pool, data discovery, external tables and views, data ingestion and transformation, Synapse pipelines, Spark pool, Power BI, Synapse Link and dedicated SQL pool.
• Most sections start with a one or two-minute overview.
• Learning tasks and assignments are spread through the course, so you practise each section on the project before moving on.

How to study it step by step

1. Watch the introduction and the warehouse versus lake videos, then create your own Synapse workspace as shown.
2. Tour each Studio hub so you know where data, scripts, pipelines and monitoring live.
3. Load the project data to the data lake and study the medallion architecture video.
4. Watch the serverless SQL pool cost control lesson early, then work through every file format: CSV, JSON, Parquet and Delta.
5. Do the data discovery learning task yourself before watching the walkthrough.
6. Build external tables and views, then complete the assignments.
7. Transform to Parquet and build the partitioned output, using stored procedures.
8. Automate it with Synapse pipelines, including parameters, dependencies and triggers.
9. Finish with Spark, Power BI, Synapse Link and the dedicated SQL pool. Remove the resources you no longer need when you are done, so they do not keep costing money.

At six to eight videos a day, plan for three to four weeks with practice. The full course is also a YouTube playlist.

If you want more practice with orchestration, the Azure Data Factory course builds a full ADF project, and Data Engineering with Microsoft Fabric covers the same kind of work in Microsoft Fabric.

Frequently asked questions

Is this Azure Synapse course free?

Yes. All 120 videos are free on YouTube. You can follow them from the course page or the playlist.

How long does it take?

The videos total about 11.6 hours. With the learning tasks and assignments, most learners need three to four weeks of regular practice.

Do I need an Azure subscription?

To build the project yourself, yes, because you create a Synapse workspace, a Spark pool, Cosmos DB and a dedicated SQL pool. The course includes a lesson on serverless SQL pool cost control; keep an eye on your spend and delete resources after practice.

Does it cover both serverless and dedicated SQL pools?

Yes. Most of the course uses the serverless SQL pool to query files in the data lake. The final section covers the dedicated SQL pool, including loading data with PolyBase and the COPY command.