MLOps with TensorFlow Extended and Apache Beam

Speaker(s): Hannes Hapke

In this talk, Hannes is providing insights into using TensorFlow Extended (TFX) and Apache Beam for MLOps. He introduces how TFX is using Apache Beam for data pipeline tasks and for orchestration entire ML pipelines. The audience learned how to run ML production pipelines with Apache Beam, and therefore, free the data scientist’s time from maintaining production machine learning models.

Hannes shows real-life examples for MLOps using TFX and Apache Beam. The talk is an introduction which is followed by a 1-hour workshop on how to write custom TFX components with TFX and Apache Beam.

Hannes is the co-author of the upcoming O’Reilly publication “Building Machine Learning Pipeline” which provides an introduction to Apache Beam for its readers. He is currently a senior data scientist at SAP Concur