Ingredients for using R as part of a production microservice ecosystem

While R is widely regarded as a primary tool for doing data science, using it in production environments as part of a wider microservice ecosystem has its challenges. At Emarsys we have used R extensively for interactive analysis and Shiny apps for three years before deciding to use R in production as well. The talk will cover why it took relatively long to do so and why it was a great decision. I will discuss the concrete use case (Bayesian modelling), the technical setup (with Kubernetes, OpenCPU, authentication, logging and continuous integration) and our lessons of how to convince a developer organization to accept R as a first-class member of its technology stack.

Pál Jenő
Data Scientist, Emarsys

Jenő has been working as a data scientist at Emarsys while also teaching courses on predictive modelling at the Central European University. 
Before that he earned a PhD in Economics at CEU and he spent two summers at Microsoft Research as an intern.