Reactive machine learning is a set of techniques for building production-grade machine learning systems that stay responsive in the face of failure or changes in load. Drawing on principles outlined in the Reactive Manifesto, the reactive approach to machine learning gives us a principled way of building large scale machine learning systems that are every bit as good as modern web apps. This talk will explore how reactive machine learning systems can be built on the JVM. We’ll look at languages like Scala and Clojure, libraries like Spark and Akka, and various reactive programming techniques like futures and actors. Then, we’ll see how we can go beyond the JVM to interact with innovative techniques implemented in other runtimes, such as deep learning in Python. The examples will be taken from real-world use cases such as fraud detection, autonomous cars, and intelligent agents. But they will be transformed into the exciting world of cartoon animals who do machine learning. Kangaroos, turtles, rabbits, and more will guide us through the wild world of reactive machine learning. Jeff Smith builds large-scale artificial intelligence systems. For the past decade, he has been working on data science applications at various startups in New York, San Francisco, and Hong Kong. Now, he leads the data engineering team behind Amy, the artificial intelligence who schedules meetings at x.ai. He is a frequent speaker, blogger, and the author of "Reactive Machine Learning Systems":https://www.manning.com/books/reactive-machine-learning-systems , an upcoming book on how to build real world machine learning systems using Scala, Akka, and Spark. [SUS-5959]