The machine learning platform must also be able to deploy analytics results, making them quickly available to consumers. If the pipeline is continuous integration for data science, deploy results can be continuous deployment. This is accomplished using Seldon, TF Serving, Clipper and OpenVINO.