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

• Introducing Apache Airflow® 3 with Vikram Koka and Jed Cunningham

  The Airflow 3.0 release marks a significant leap forward in modern data orchestration, introducing architectural upgrades that improve scalability, flexibility and long-term maintainability.In this episode, we welcome Vikram Koka, Chief Strategy Officer at Astronomer, and Jed Cunningham, Principal Software Engineer at Astronomer, to discuss the architectural foundations, new features and future implications of this milestone release. They unpack the rationale behind DAG versioning and task execution interface, explain how Airflow now integrates more seamlessly within broader data ecosystems and share how these changes lay the groundwork for multi-cloud deployments, language-agnostic workflows and stronger enterprise security.Key Takeaways:(02:28) Modern orchestration demands new infrastructure approaches.(05:02) Removing legacy components strengthens system stability.(06:26) Major releases provide the opportunity to reduce technical debt.(08:31) Frontend and API modernization enable long-term adaptability.(09:36) Event-based triggers expand integration possibilities.(11:54) Version control improves visibility and execution reliability.(14:57) Centralized access to workflow definitions increases flexibility.(21:49) Decoupled architecture supports distributed and secure deployments.(26:17) Community collaboration is essential for sustainable growth.Resources Mentioned:Astronomer Websitehttps://www.astronomer.ioApache Airflowhttps://airflow.apache.org/Git Bundlehttps://git-scm.com/book/en/v2/Git-Tools-BundlingFastAPIhttps://fastapi.tiangolo.com/Reacthttps://react.dev/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

  --------  

  27:28
• Airflow in Action: Powering Instacart's Complex Ecosystem

  The evolution of data orchestration at Instacart highlights the journey from fragmented systems to robust, standardized infrastructure. This transformation has enabled scalability, reliability and democratization of tools for diverse user personas.In this episode, we’re joined by Anant Agarwal, Software Engineer at Instacart, who shares insights into Instacart's Airflow journey, from its early adoption in 2019 to the present-day centralized cluster approach. Anant discusses the challenges of managing disparate clusters, the implementation of remote executors, and the strategic standardization of infrastructure and DAG patterns to streamline workflows.Key Takeaways:(03:49) The impact of external events on business growth and technological evolution.(04:31) Challenges of managing decentralized systems across multiple teams.(06:14) The importance of standardizing infrastructure and processes for scalability.(09:51) Strategies for implementing efficient and repeatable deployment practices.(12:17) Addressing diverse user personas with tailored solutions.(14:47) Leveraging remote execution to enhance flexibility and scalability.(18:36) Benefits of transitioning to a centralized system for organization-wide use.(20:57) Maintaining an upgrade cadence to stay aligned with the latest advancements.(23:35) Anticipation for new features and improvements in upcoming software versions.Resources Mentioned:Anant Agarwalhttps://www.linkedin.com/in/anantag/Instacart | LinkedInhttps://www.linkedin.com/company/instacart/Instacart | Websitehttps://www.instacart.comApache Airflowhttps://airflow.apache.org/AWS Amazonhttps://aws.amazon.com/ecs/Terraformhttps://www.terraform.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 Airflow for Data Engineering & 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

  --------  

  25:14
• From ETL to Airflow: Transforming Data Engineering at Deloitte Digital with Raviteja Tholupunoori

  Data orchestration at scale presents unique challenges, especially when aiming for flexibility and efficiency across cloud environments. Choosing the right tools and frameworks can make all the difference. In this episode, Raviteja Tholupunoori, Senior Engineer at Deloitte Digital, joins us to explore how Airflow enhances orchestration, scalability and cost efficiency in enterprise data workflows.Key Takeaways:(01:45) Early challenges in data orchestration before implementing Airflow.(02:42) Comparing Airflow with ETL tools like Talend and why flexibility matters.(04:24) The role of Airflow in enabling cloud-agnostic data processing.(05:45) Key lessons from managing dynamic DAGs at scale.(13:15) How hybrid executors improve performance and efficiency.(14:13) Best practices for testing and monitoring workflows with Airflow.(15:13) The importance of mocking mechanisms when testing DAGs.(17:57) How Prometheus, Grafana and Loki support Airflow monitoring.(22:03) Cost considerations when running Airflow on self-managed infrastructure.(23:14) Airflow’s latest features, including hybrid executors and dark mode.Resources Mentioned:Raviteja Tholupunoorihttps://www.linkedin.com/in/raviteja0096/?originalSubdomain=inDeloitte Digitalhttps://www.linkedin.com/company/deloitte-digital/Apache Airflowhttps://airflow.apache.org/Grafanahttps://grafana.com/solutions/apache-airflow/monitor/Astronomer Presents: Exploring Apache Airflow® 3 Roadshowshttps://www.astronomer.io/events/roadshow/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 Airflow for Data Engineering & 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

  --------  

  27:42
• A Deep Dive Into the 2025 State of Airflow Survey Results with Tamara Fingerlin of Astronomer

  The 2025 State of Airflow report sheds light on how global users are adopting, evolving and innovating with Apache Airflow. With over 5,000 responses from 116 countries, the survey reveals critical insights into Airflows’ role in business operations, new use cases and what’s ahead for the community.In this episode, Tamara Fingerlin, Developer Advocate at Astronomer, walks us through her process of analyzing survey data, key trends from the report and what to expect from Airflow 3.0.Key Takeaways:(02:14) The State of Airflow report combines anonymized telemetry and survey results.(03:25) The survey received thousands of responses from many countries, showcasing global reach.(04:49) The survey process involves multiple steps, from question selection to report creation.(09:00) Many users expect to increase Airflow usage for revenue-generating or external use cases.(11:04) Experienced users tend to utilize Airflow more for advanced use cases like MLOps.(15:13) UI improvements offer enhanced navigation and error visibility.(18:15) Architectural changes enable new capabilities like remote execution and language support.(19:40) Long-requested features will be available in the new major release.(21:00) Future aspirations include integrating data visualization capabilities into the UI.Resources Mentioned:Tamara Fingerlinhttps://www.linkedin.com/in/tamara-janina-fingerlin/Astronomer | LinkedIn https://www.linkedin.com/company/astronomer/Astronomer | Websitehttps://www.astronomer.ioApache Airflowhttps://airflow.apache.org/2025 State of Airflow Webinarhttps://www.astronomer.io/airflow/state-of-airflow/Airflow Slackhttps://apache-airflow-slack.herokuapp.com/Astronomer Presents: Exploring Apache Airflow® 3 Roadshowshttps://www.astronomer.io/events/roadshow/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 Airflow for Data Engineering & 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

  --------  

  23:26
• Airflow’s Role in the Rise of DataOps with Andy Byron

  The orchestration layer is evolving into a critical component of the modern data stack. Understanding its role in DataOps is key to optimizing workflows, improving reliability and reducing complexity.In this episode, Andy Byron, CEO at Astronomer, discusses the rapid growth of Apache Airflow, the increasing importance of orchestration and how Astronomer is shaping the future of DataOps.Key Takeaways:(01:54) Orchestration is central to modern data workflows.(03:16) Airflow 3.0 will enhance usability and flexibility.(05:14) AI-driven workloads demand zero-downtime orchestration.(08:13) DataOps relies on orchestration for seamless operations.(11:05) Integration across ingestion, transformation and governance is key.(17:24) The future of DataOps is consolidation and automation.(19:13) Enterprises use Airflow to process massive data volumes.(23:20) Product innovation is driven by customer needs and feedback.Resources Mentioned:Andy Byronhttps://www.linkedin.com/in/andy-byron-417a429/Astronomer | LinkedInhttps://www.linkedin.com/company/astronomer/Astronomer | Websitehttps://www.astronomer.ioApache Airflowhttps://airflow.apache.org/State of Airflow Webinarhttps://www.astronomer.io/events/webinars/the-state-of-airflow-2025-video/Astronomer Observehttps://www.astronomer.io/product/observe/Astronomer Roadshow: Exploring Apache Airflow 3 | Londonhttps://www.astronomer.io/events/roadshow/london/Astronomer Roadshow: Exploring Apache Airflow 3 | New Yorkhttps://www.astronomer.io/events/roadshow/new-york/Astronomer Roadshow: Exploring Apache Airflow 3 | Sydneyhttps://www.astronomer.io/events/roadshow/sydney/Astronomer Roadshow: Exploring Apache Airflow 3 | San Franciscohttps://www.astronomer.io/events/roadshow/san-francisco/Astronomer Roadshow: Exploring Apache Airflow 3 | Chicagohttps://www.astronomer.io/events/roadshow/chicago/Thanks for listening to “The Data Flowcast: Mastering Airflow for Data Engineering & 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

  --------  

  26:15

Plus de podcasts Technologies

Podcasts tendance de Technologies

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

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