I have a project written in R with multiple R files and I want to deploy it on Kubernetes for model inference.

Some articles I can find relevant to this is to submit the project to a k8s cluster through RStudio. This is good for model training but not applicable in model inference, unless there is a way to launch the R project on kubernetes without using RStudio.

https://blog.rstudio.com/2018/10/01/sparklyr-0-9/

https://aws.amazon.com/blogs/big-data/running-sparklyr-rstudios-r-interface-to-spark-on-amazon-emr/

https://github.com/sparklyr/sparklyr/issues/1525

What is the recommended way to serve the models written using Sparklyr?

Ideally I would like to deploy it in K8s. AWS EMR works too. Maybe there is a way to use Databricks notebook to serve the R models?