I have a simple data pipeline where the data and the functional code are stored in a custom R package hosted on Azure DevOps and the final outputs are produced by quarto documents pointing to specific branches of the package.