Jan will talk about his experience building several production-grade systems that heavily rely on machine learning & computer vision.
He will talk about the challenges in not only the machine learning (and deep learning) code, but also about test and validation data management, the development & experimentation environments; then on the “production” side, model storage & serving, validation data reporting, and obviously continuous integration & deployment.
There will be plenty of code examples throughout, both in the ML as well as the infrastructure arena; possibly even a daring demo!
People who have heard enough about machine learning, infrastructure as code, and distributed systems to be dangerous, but who are keen to learn the details. Also people who are building a ML system, but aren’t quite sure how a production ML system looks like under the hood.
Keyword: Scala, Akka, Deep Learning
Session Category : Talk