If you’re a data driven organization, it’s likely that you’ve considered building a machine learning application.Generally speaking, machine learning can enable your team to do anything from give smart recommendations to your users in your product, or predict user growth based on past behavior. In order for that machine learning model to result in that output for your team, however, it requires actually operationalizing it. It’s one thing to write a data science pipeline, and it’s another to have it running in production, reliably, on a schedule.
This conference talk will give the DataConnect audience a glimpse and stronger understanding of what it takes to do MLOps, and why you need data orchestration to do it well. Specifically, we’ll cover:
● What is MLOps, and how is it different than Machine Learning? What are the business problems that we’re typically solving for?
● What does the profile of the person who does MLOps look like, and how has that changed over the last few years?
● What does a typical end-to-end machine learning pipeline look like, why do you need to operationalize it, and what challenges might you hit along the way?
● What’s the toolset typically at your disposal to solve that problem?
● What’s the role of data orchestration and Apache Airflow in that process?
● Example ML pipeline with Airflow
● Reflections on Astronomer + the future
Paola is a product leader and proud co-founder at Astronomer, the company behind the Apache Airflow open source project. Astronomer's goal is to make it easier for data engineers to write and run data pipelines. Paola has spent her 5+ years at Astronomer wearing many hats, but her core is in product management and developer education. She’s made most of her impact by curating developer experiences in Astronomer’s cloud service and leading an Education and Docs team that helps folks learn Astro and Airflow. Paola is a graduate of Georgetown University and spent 5 years in Cincinnati, Ohio, as a Venture for America Fellow. Originally from Mexico City, she currently lives in Brooklyn and is excited to keep growing Astronomer’s footprint.
Paola is a product leader and proud co-founder at Astronomer, the company behind the Apache Airflow open source project. Astronomer's goal is to make it easier for data engineers to write and run data pipelines. Paola has spent her 5+ years at Astronomer wearing many hats, but her core is in product management and developer education. She’s made most of her impact by curating developer experiences in Astronomer’s cloud service and leading an Education and Docs team that helps folks learn Astro and Airflow. Paola is a graduate of Georgetown University and spent 5 years in Cincinnati, Ohio, as a Venture for America Fellow. Originally from Mexico City, she currently lives in Brooklyn and is excited to keep growing Astronomer’s footprint.