Machine Learning Deployment Using Docker

Abstract

Over the last few years, Docker has changed the way applications are deployed in production. Application architectures have moved from monolithic to microservices, with more control of continuous (ongoing) deployments that don’t impact a large of part of the running applications. Docker has proved to be instrumental in allowing applications to run at scale and to be available all the time. Though it’s been more than seven years since Docker was released, it’s gotten a lot of attention from the developer community recently (especially by DevOps and MLOps teams). Companies large and small are using Docker in applications.