Kenten Danas

Apache Airflow® has long been the control plane for data pipelines. As AI workflows move into production, teams are discovering the same challenges apply: LLM calls fail, embeddings need regenerating, and agent outputs need human review.

In a high-pressure kitchen, great service depends on mise en place — everything prepared, organized, and ready before the rush. Production AI demands the same discipline. This talk shows how Apache Airflow unifies data pipelines and AI workflows in one observable control plane and system of record — replacing fragmented observability, failure handling, and team handoffs. We'll walk through real projects covering multi-agent orchestration, AI-driven pipeline branching, and vision model integration — each illustrating a distinct production pattern and all running on Airflow. The best kitchens don't just execute recipes — they make the rush look effortless.

Airflow 3.0 is the most significant release in the project’s history, and brings a better user experience, stronger security, and the ability to run tasks anywhere, at any time.

Airflow 3 transforms data orchestration with three game-changing capabilities: - Intuitive experience: Revamped UI, DAG versioning and integrated backfills streamline pipeline management - Enterprise security: Task isolation architecture eliminates direct database connections, reducing exposure risks - Boundless flexibility: Execute tasks anywhere — from edge to hybrid-cloud to GPU workloads — with remote execution Discover how these innovations are already revolutionizing how forward-thinking teams operationalize mission-critical data workflows.

Join us for a two-hour hands-on workshop to kickstart your journey with Apache Airflow. Trusted by millions, Airflow is the go-to solution for orchestrating complex data pipelines.

Machine Learning models can add value and insight to many projects, but they can be challenging to put into production due to problems like lack of reproducibility, difficulty maintaining integrations, and sneaky data quality issues.