The Data Flowcast: Mastering Apache Airflow ® for Data Engineering and AI

Managing modern data platforms means navigating a web of complex infrastructure, competing team needs and evolving security standards. For data teams to truly thrive, infrastructure must become both accessible and compliant without sacrificing velocity or reliability.

In this episode, we’re joined by Cory O’Daniel, CEO and Co-Founder at Massdriver, and Jacob Ferriero, Senior Software Engineer at Astronomer, to unpack what it takes to make data platform engineering scalable, sustainable and secure. They share lessons from years of experience working with DevOps, ML teams and platform engineers and discuss how Airflow fits into the orchestration layer of today’s data stacks.

Key Takeaways:

(03:27) Making infrastructure accessible without deep ops knowledge.

(07:23) Distinct personas and responsibilities across data teams.

(09:53) Infrastructure hurdles specific to ML workloads.

(11:13) Compliance and governance shaping platform design.

(13:27) Tooling mismatches between teams cause friction.

(15:13) Airflow’s orchestration role within broader system architecture.

(22:10) Creating reusable infrastructure patterns for consistency.

(24:13) Enabling secure access without slowing down development.

(26:55) Opportunities to improve Airflow with event-driven and reliability tooling.

Resources Mentioned:

Cory O’Daniel

https://www.linkedin.com/in/coryodaniel/

Massdriver | LinkedIn

https://www.linkedin.com/company/massdriver/

Massdriver | Website

https://www.massdriver.cloud/

Jacob Ferriero

https://www.linkedin.com/in/jacob-ferriero/

Astronomer

https://www.linkedin.com/company/astronomer/

Apache Airflow

https://airflow.apache.org/

Prequel

https://www.prequel.co/

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

Telemetry has the potential to guide the future of Airflow, but only if it’s implemented transparently and with community trust. 

In this episode, we’re joined by Bolke de Bruin, Director at Metyis and a long-time Airflow PMC member. Bolke discusses how telemetry has been handled in the past, why it matters now and what it will take to get it right.

Key Takeaways:

(03:20) The role of foundations in establishing credibility and sustainability.

(04:52) Why data collection is critical to open-source project direction.

(07:24) Lessons learned from previous approaches to user data collection.

(10:23) The current state of telemetry in the project.

(10:53) Community trust as a prerequisite for technical implementation.

(12:54) The importance of managing sensitive data within trusted ecosystems.

(16:37) Ethical considerations in balancing participation and access.

(18:45) Forward-looking ideas for improving workflow design and usability.

Resources Mentioned:

Bolke de Bruin

https://www.linkedin.com/in/bolke/

Metyis | LinkedIn

https://www.linkedin.com/company/metyis/

Metyis | Website

http://www.metyis.com

Apache Airflow

https://airflow.apache.org/

Airflow Summit

https://airflowsummit.org/

Airflow Dev List

https://lists.apache.org/[email protected]

https://www.astronomer.io/events/roadshow/london/

https://www.astronomer.io/events/roadshow/new-york/ 

https://www.astronomer.io/events/roadshow/sydney/   

https://www.astronomer.io/events/roadshow/san-francisco/  

https://www.astronomer.io/events/roadshow/chicago/ 

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

Contributing to open-source projects can be daunting, but it can also unlock unexpected innovation. This episode showcases how one engineer’s journey with Apache Airflow led to impactful UI enhancements and infrastructure solutions at scale. Shubham Raj, Software Engineer II at Cloudera, shares how his team built a drag-and-drop DAG editor for non-coders, contributions which helped shape the Airflow 3.0 Ul and introduced features like external XCom control and bulk APls.



Key Takeaways:

(02:30) Day-to-day responsibilities building platforms that simplify orchestration.

(05:27) Factors that make onboarding into large open-source projects accessible.

(07:35) The value of improved user interfaces for task state visibility and control.

(09:49) Enabling faster debugging by exposing internal data through APIs.

(13:00) Balancing frontend design goals with backend functionality.

(14:19) Creating workflow editors that lower the barrier to entry.

(16:54) Supporting a variety of task types within a visual DAG builder.

(19:32) Common infrastructure challenges faced by orchestration users.

(20:37) Addressing dependency management across distributed environments.

Resources Mentioned:

Shubham Raj

https://www.linkedin.com/in/shubhamrajofficial/

Cloudera | LinkedIn

https://www.linkedin.com/company/cloudera/

Cloudera | Website

https://www.cloudera.com/

Apache Airflow

https://airflow.apache.org/

2023 Airflow Summit

https://airflowsummit.org/

https://www.astronomer.io/events/roadshow/london/  

https://www.astronomer.io/events/roadshow/new-york/  

https://www.astronomer.io/events/roadshow/sydney/  

https://www.astronomer.io/events/roadshow/san-francisco/  

https://www.astronomer.io/events/roadshow/chicago/

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

Managing data pipelines at scale is not just a technical challenge. It is also an organizational one. At Lyft, success means empowering dozens of teams to build with autonomy while enforcing governance and best practices across thousands of workflows.

In this episode, we speak with Yunhao Qing, Software Engineer at Lyft, about building a governed data-engineering platform powered by Airflow that balances flexibility, standardization and scale.

Key Takeaways:

(03:17) Supporting internal teams with a centralized orchestration platform.

(04:54) Migrating to a managed service to reduce infrastructure overhead.

(06:04) Embedding platform-level governance into custom components.

(08:02) Consolidating and regulating the creation of custom code.

(09:48) Identifying and correcting inefficient workflow patterns.

(11:17) Replacing manual workarounds with native platform features.

(14:32) Preparing teams for major version upgrades.

(16:03) Leveraging asset-based scheduling for smarter triggers.

(18:13) Envisioning GenAI and semantic search for future productivity.

Resources Mentioned:

Yunhao Qing

https://www.linkedin.com/in/yunhao-qing

Lyft | LinkedIn

https://www.linkedin.com/company/lyft/

Lyft | Website

https://www.lyft.com/

Apache Airflow

https://airflow.apache.org/

Astronomer

https://www.astronomer.io/

Kubernetes

https://kubernetes.io/

https://www.astronomer.io/events/roadshow/london/

https://www.astronomer.io/events/roadshow/new-york/  

https://www.astronomer.io/events/roadshow/sydney/  

https://www.astronomer.io/events/roadshow/san-francisco/  

https://www.astronomer.io/events/roadshow/chicago/

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

Understanding the complexities of Apache Airflow can be daunting for newcomers and seasoned data engineers. But with the right guidance, mastering the tool becomes an achievable milestone.

In this episode, Marc Lamberti, Head of Customer Education at Astronomer, joins us to share his journey from Udemy instructor to driving education at Astronomer, and how he's helping over 100,000 learners demystify Airflow.

Key Takeaways:

(02:36) Early exposure to Airflow while addressing inefficiencies in data workflows.

(04:10) Common barriers to implementing open source tools in enterprise settings.

(06:18) The shift from part-time teaching to a full-time focus on Airflow education.

(07:53) A modular, guided approach to structuring educational content.

(09:57) The value of highlighting underused Airflow features for broader adoption.

(12:35) Certifications as a method to assess readiness and uncover knowledge gaps.

(13:25) Coverage of essential Airflow concepts in the Fundamentals exam.

(16:07) The DAG Authoring exam’s emphasis on practical, advanced features.

(20:08) A call for more visible integration of Airflow with AI workflows.

Resources Mentioned:

Marc Lamberti

https://www.linkedin.com/in/marclamberti/

Astronomer | LinkedIn

https://www.linkedin.com/company/astronomer/

Astronomer Academy

https://academy.astronomer.io/

Airflow Fundamentals Certification

https://www.astronomer.io/certification/

DAG Authoring Certification

https://academy.astronomer.io/plan/astronomer-certification-dag-authoring-for-apache-airflow-exam

The Complete Hands-On Introduction to Airflow

https://www.udemy.com/course/the-complete-hands-on-course-to-master-apache-airflow/?utm_source=adwords&utm_medium=udemyads&utm_campaign=Search_DSA_Beta_Prof_la.EN_cc.ROW-English&campaigntype=Search&portfolio=ROW-English&language=EN&product=Course&test=&audience=DSA&topic=&priority=Beta&utm_content=deal4584&utm_term=_._ag_162511579404_._ad_696197165418_._kw__._de_c_._dm__._pl__._ti_dsa-1677053911088_._li_9061346_._pd__._&matchtype=&gad_source=1&gad_campaignid=21168154305&gbraid=0AAAAADROdO3MpljfP-gssiYSmDEPdhZV9&gclid=Cj0KCQjw097CBhDIARIsAJ3-nxdjZA6G5-Y0-akk6Huksy2PLb04t92J4iNfUSIbMdrSAla_tb-o2N8aArOeEALw_wcB&couponCode=PMNVD3025

https://www.astronomer.io/events/roadshow/london/  

https://www.astronomer.io/events/roadshow/new-york/  

https://www.astronomer.io/events/roadshow/sydney/  

https://www.astronomer.io/events/roadshow/san-francisco/  

https://www.astronomer.io/events/roadshow/chicago/

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

The flexibility of Airflow plays a pivotal role in enabling decentralized data architectures and empowering cross-functional teams.

In this episode, we speak with Alberto Crespi, Data Architect at lastminute.com, who shares how his team scales Airflow across 12 teams while supporting both vertical and horizontal structures under a data mesh approach.

Key Takeaways:

(02:17) Defining responsibilities within data architecture teams.

(04:15) Consolidating multiple orchestrators into a single solution.

(07:00) Scaling Airflow environments with shared infrastructure and DevOps practices.

(10:59) Managing dependencies and readiness using SQL sensors.

(14:23) Enhancing visibility and response through Slack-integrated monitoring.

(19:28) Extending Airflow’s flexibility to run legacy systems.

(22:28) Integrating transformation tools into orchestrated pipelines.

(25:54) Enabling non-engineers to contribute to pipeline development.

(27:33) Fostering adoption through collaboration and communication.

Resources Mentioned:

Alberto Crespi

https://www.linkedin.com/in/crespialberto/

lastminute.com | Website

https://lastminute.com

Apache Airflow

https://airflow.apache.org/

dbt Labs

https://www.getdbt.com/

Astronomer Cosmos

https://github.com/astronomer/astronomer-cosmos

GitLabSlack

https://slack.com/

Kubernetes

https://kubernetes.io/

Confluence

https://www.atlassian.com/software/confluence

Slack

https://slack.com/

https://www.astronomer.io/events/roadshow/london/

https://www.astronomer.io/events/roadshow/new-york/

https://www.astronomer.io/events/roadshow/sydney/

https://www.astronomer.io/events/roadshow/san-francisco/

https://www.astronomer.io/events/roadshow/chicago/

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

Innovation in orchestration is redefining how engineers approach both traditional ETL pipelines and emerging AI workloads. Understanding how to harness Airflow’s flexibility and observability is essential for teams navigating today’s evolving data landscape.

In this episode, Anu Pabla, Principal Engineer at The ODP Corporation, joins us to discuss her journey from legacy orchestration patterns to AI-native pipelines and why she sees Airflow as the future of AI workload orchestration.

Key Takeaways:

(03:43) Engaging with external technology communities fosters innovation.

(05:05) Mentoring early-career engineers builds confidence in a complex tech landscape.

(07:51) Orchestration patterns continue to evolve with modern data needs.

(08:41) Managing AI workflows requires structured and flexible orchestration.

(10:35) High-quality, meaningful data remains foundational across use cases.

(15:08) Community-driven open source tools offer lasting value.

(16:59) Self-healing systems support both legacy and AI pipelines.

(20:20) Orchestration platforms can drive future AI-native workloads.

Resources Mentioned:

Anu Pabla

https://www.linkedin.com/in/atomicap/

The ODP Corporation

https://www.linkedin.com/company/the-odp-corporation/

The ODP Corporation | Website

https://www.theodpcorp.com/homepage

Apache Airflow

https://airflow.apache.org/

LlamaIndex

https://www.llamaindex.ai/

https://www.astronomer.io/events/roadshow/london/  

https://www.astronomer.io/events/roadshow/new-york/  

https://www.astronomer.io/events/roadshow/sydney/  

https://www.astronomer.io/events/roadshow/san-francisco/  

https://www.astronomer.io/events/roadshow/chicago/

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

The orchestration layer is foundational to building robust AI- and ML-powered data pipelines, especially in complex hybrid enterprise environments. IBM’s partnership with Astronomer reflects a strategic alignment to simplify and scale Airflow-based workflows across industries.

In this episode, we’re joined by IBM’s Senior Product Manager, BJ Adesoji, and GTM PM and Growth Leader, Ryan Yackel. We discuss how IBM customers are using Airflow in production, the challenges they face at scale and what the new IBM–Astronomer collaboration unlocks.

Key Takeaways:

(03:09) The growing importance of orchestration tools in enterprise environments.

(04:48) How organizations are expanding orchestration beyond traditional use cases.

(05:24) Common patterns across industries adopting orchestration platforms.

(07:16) Why orchestration is essential for supporting business-critical workloads.

(10:00) The role of orchestration in compliance and regulatory processes.

(13:02) Challenges enterprises face when managing orchestration infrastructure.

(14:58) Opportunities to simplify and centralize orchestration at scale.

(19:11) The value of integrating orchestration with broader data toolchains.

(20:54) How AI is shaping the future of orchestrated data workflows.

Resources Mentioned:

BJ Adesoji

https://www.linkedin.com/in/bj-soji/

Ryan Yackel

https://www.linkedin.com/in/ryanyackel/

IBM | LinkedIn

https://www.linkedin.com/company/databand-ai/

IBM Databand

https://www.ibm.com/products/databand

IBM DataStage

https://www.ibm.com/products/datastage

IBM watsonx.governance

https://www.ibm.com/products/watsonx-governance

IBM Knowledge Catalog

https://www.ibm.com/products/knowledge-catalog

Apache Airflow

https://airflow.apache.org/

watsonx Orchestrate

https://www.ibm.com/products/watsonx-orchestrate

Domino

https://domino.ai/

Astronomer

https://www.astronomer.io/

Snowflake

https://www.snowflake.com/en/

dbt Labs

https://www.getdbt.com/

Amazon SageMaker

https://aws.amazon.com/sagemaker/

Cloudera

https://www.cloudera.com/

MongoDB

https://www.mongodb.com/

https://www.astronomer.io/events/roadshow/london/ 

https://www.astronomer.io/events/roadshow/new-york/ 

https://www.astronomer.io/events/roadshow/sydney/ 

https://www.astronomer.io/events/roadshow/san-francisco/ 

https://www.astronomer.io/events/roadshow/chicago/ 

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

Efficient orchestration and maintainability are crucial for data engineering at scale. Gil Reich, Data Developer for Data Science at Wix, shares how his team reduced code duplication, standardized pipelines, and improved Airflow task orchestration using a Python-based framework built within the data science team.

In this episode, Gil explains how this internal framework simplifies DAG creation, improves documentation accuracy, and enables consistent task generation for machine learning pipelines. He also shares lessons from complex DAG optimization and maintaining testable code.

Key Takeaways:

(03:23) Code duplication creates long-term problems.

(08:16) Frameworks bring order to complex pipelines.

(09:41) Shared functions cut down repetitive code.

(17:18) Auto-generated docs stay accurate by design.

(22:40) On-demand DAGs support real-time workflows.

(25:08) Task-level sensors improve run efficiency.

(27:40) Combine local runs with automated tests.

(30:09) Clean code helps teams scale faster.

Resources Mentioned:

Gil Reich

https://www.linkedin.com/in/gilreich/

Wix | LinkedIn

https://www.linkedin.com/company/wix-com/

Wix | Website

https://www.wix.com/

DS DAG Framework

https://airflowsummit.org/slides/2024/92-refactoring-dags.pdf

Apache Airflow

https://airflow.apache.org/

https://www.astronomer.io/events/roadshow/london/  

https://www.astronomer.io/events/roadshow/new-york/  

https://www.astronomer.io/events/roadshow/sydney/  

https://www.astronomer.io/events/roadshow/san-francisco/  

https://www.astronomer.io/events/roadshow/chicago/

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning

Legacy architecture and AI workloads pose unique challenges at scale, especially in a global enterprise with complex data systems. In this episode, we explore strategies to proactively monitor and optimize pipelines while minimizing downstream failures.

Adonis Castillo Cordero, Senior Automation Manager at Procter & Gamble, joins us to share actionable best practices for dependency mapping, anomaly detection and architecture simplification using Apache Airflow.

Key Takeaways:

(03:13) Integrating legacy data systems into modern architecture.

(05:51) Designing workflows for real-time data processing.

(07:57) Mapping dependencies early to avoid pipeline failures.

(09:02) Building automated monitoring into orchestration frameworks.

(12:09) Detecting anomalies to prevent performance bottlenecks.

(15:24) Monitoring data quality to catch silent failures.

(17:02) Prioritizing responses based on impact severity.

(18:55) Simplifying dashboards to highlight critical metrics.

Resources Mentioned:

Adonis Castillo Cordero

https://www.linkedin.com/in/adoniscc/

Procter & Gamble | LinkedIn

https://www.linkedin.com/company/procter-and-gamble/

Procter & Gamble | Website

http://www.pg.com

Apache Airflow

https://airflow.apache.org/

OpenLineage

https://openlineage.io/

Azure Monitor

https://azure.microsoft.com/en-us/products/monitor/

AWS Lookout for Metrics

https://aws.amazon.com/lookout-for-metrics/

Monte Carlo

https://www.montecarlodata.com/

Great Expectations

https://greatexpectations.io/

https://www.astronomer.io/events/roadshow/london/

https://www.astronomer.io/events/roadshow/new-york/

https://www.astronomer.io/events/roadshow/sydney/ 

https://www.astronomer.io/events/roadshow/san-francisco/ 

https://www.astronomer.io/events/roadshow/chicago/

Thanks for listening to “The Data Flowcast: Mastering Apache Airflow® for Data Engineering and AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.

#AI #Automation #Airflow #MachineLearning