MLOps For Python with R code

Question

Hello Azure MLOps Team,

I am looking at the reference architecture https://learn.microsoft.com/en-us/azure/architecture/reference-architectures/ai/mlops-python.

Most of our data science models are made in R. I am wondering, which part of the MLOps process (Azure Pipelines or Azure ML Compute/Azure ML Pipelines) can be configured in Python for R code to run?

Our preference is to leverage as much Python as possible for the code R code. I see there is an Azure ML SDK for python and R. Can we use Azure ML SDK in Python with R codebase?

Kind regards,
Slava Keshkov

Answer 1

Hi Slava,

I found this for you: https://rstudio.com/resources/rstudioconf-2020/mlops-for-r-with-azure-machine-learning/

In the video David is talking about Devops for R at 10:58. Why dont you ping David on Twitter to inquire more about it? If Microsoft is not able to support it officially I think you can build your own pipelines deploying models within AKS and using Azure storage with Azure Functions for pushing data to the model and pulling out forcasts.

Hope this helps.

Thanks,
Vaibhav

Answer 2

Here's the azureml sdk for R repo for your reference.