Permission to implement RAG on AZUR Databricks

Hello @Karishma Nanda Welcome to Q&A forum! The process of deploying RAG models on Azure Databricks is similar to deploying other machine learning models on Azure Databricks. You can use the same deployment process and tools that you would use for other machine learning models. Please try the below permissions and do let us know!

1. Contributor or Owner role on the Azure Databricks workspace resource.
2. Contributor or Owner role on the Azure Machine Learning workspace resource, if you are using Azure Machine Learning to deploy the RAG model.

In addition to these permissions, you will also need to have the necessary access and credentials to the data sources that you are using to train and deploy the RAG model. Region availability The features that support RAG application development on Databricks are available in the same regions as model serving. If you plan on using Foundation Model APIs as part of your RAG application development, you are limited to the supported regions for Foundation Model APIs. I hope this helps! Let me know if you have any further questions.

References: Link1 Link2 Link3

https://www.databricks.com/product/machine-learning/retrieval-augmented-generation