Google Cloud AI Platform offers a flexible set of tools for all steps along the machine learning journey. Although Cloud AI Platform most prominently supports Python-based workflows, it's really possible to use any language thanks to the platform's use of Docker.
This repo offers code and tutorials for how to use R with AI Platform.
- Notebooks: how to use RStudio + R in a managed notebook instance
- Training: how to specify R code to run in batch training job
- Prediction: how to serve a model created in R
Although there are some R packages for interacting with Google Cloud APIs, you'll find the best luck using gcloud in your code (through system). To call APIs not available from gcloud, you can build the respective discovery client using (googleAuthR)[https://cran.r-project.org/web/packages/googleAuthR/vignettes/building.html]