Deploying Machine Learning has never been so easy – Towards Data Science

When it comes to building Machine Learning apps for a company's needs, there are certain best practices that engineers may be keen to follow. I personally take my inspiration from Google's Rules of ML, in particular: Coming from a research-oriented academic background, I admit this is a rule I could have easily overlooked. The massive growth of Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) providers makes it easier for Machine Learning engineers to deploy live products and ensure estimator pipeline reliability. In this post, I am showing how to deploy and serve a simple Machine Learning model, leveraging cloud services. I will focus on PaaS Google App Engine, which is a fully-managed versatile tool, allowing engineers to build almost any kind of application. App Engine comes with standard runtimes in most of the popular languages (Java, PHP, Node.js,