Three features now available in preview enable you to enjoy simplified management of the customer-managed key (CMK) encrypted workspace, take advantage of our content filtering system alongside models that you can access through Models as a Service, and bring in indexes from Cosmos DB for MongoDB vCore and Pinecone for use in RAG workflows.
Service-side encryption of metadata (Simplified CMK architecture): You can now take advantage of the customer-managed key-encryption workspace’s new architecture, which will ultimately reduce costs and limit the likelihood of running into Azure policy conflicts.
Protect yourself and your users from potentially harmful content: You can now guard against potentially harmful content for you and your users, and utilize the flexibility to turn content filtering on or off based on the needs of your business.
Use RAG patterns across multiple data sources in AzureML & AI Studio: You will soon be able to bring your own indexes without having to create them from scratch, allowing you to meet customers where data is already available and giving you the flexibility to use more index types.
