Apache Airflow^® is an extremely flexible and powerful tool, capable of orchestrating almost any complex workflow touching a myriad of tools. Airflow 3, released in April 2025, is the most powerful Airflow version yet. It includes new features that improve the user experience, strengthen security, and give you more flexibility over when and where you run tasks; many of these features came directly from user requests made in previous Airflow surveys!

In this hands-on workshop, you will learn how to use essential Airflow 3 features in the context of implementing ETL/ELT workflows, all in your browser, no setup required. Work through a realistic scenario processing interplanetary travel data for AstroTrips, a fictional space tourism company.

For this workshop, you will use a free trial of Astro to run Airflow and the Astro IDE to write Dags. It is not necessary to understand all details of the Astro platform, but in a nutshell: Astro allows you to create hosted, fully managed Airflow Deployments that make running Airflow in production easy, while providing additional functionality for writing data pipelines with the Astro IDE.

What you’ll learn

• Airflow 3 exclusive features: Dag versioning, backfills, and how to use the new Airflow UI.
• Data-driven orchestration: Trigger pipelines automatically when upstream data lands with assets.
• Human-in-the-loop: How Airflow can become a user interface, pause workflows for manual review or approval.
• Other Airflow best practices including data quality checks and dynamic task mapping

Workshop Details

• Focus: Airflow 3 features for ETL/ELT patterns
• Duration: 90 minutes
• Level: Beginner to intermediate
• Prerequisites: Basic Python knowledge
• Environment: Browser-based (Astro IDE), no local setup / Docker required