R is not built for general purpose programming, but people would still like to take what they've built in R and productize it without too much context switching

R is not built for general purpose programming, but people would still like to take what they've built in R and productize it without too much context switching. Python can do this pretty well. Build and test a model with numpy, pandas, scikit-learn and then build an API around your production model with Flask and serve it on AWS or GCP. This is currently not so easy with R.