Deploying ML models via Plumber

A working example for deploying a Machine Learning model via the plumber package. This repository also contains slides used for a talk given at RStudio::conf 2019.

🎥 of RStudio::conf 2019 talk

There are two separate pieces to this example. The model-api files provide scripts for both training a simple model and building a Plumber API to serve that model. The api-app files build a simple shiny application to interact with the deployed model API. The native-app files create a separate shiny app that uses the R model but does not use the API. These two files can be used to show how an existing application (native-app) can be updated to leverage an API.