Content by khelan modi (3)
Khelan Modi introduces Microsoft Embedding 1, a multimodal model for retrieving text, images, and document pages through Azure Data AI. The article compares retrieval accuracy and storage trade-offs, provides Python and C# examples, and explains how to store and search dense vectors in Azure Cosmos DB.
Khelan Modi demonstrates Azure DocumentDB’s full-text search preview, including relevance ranking, typo tolerance, phrase matching, structured filters, and custom analyzers. The walkthrough shows how to create search indexes and query products, explains preview limitations, and outlines how to combine keyword and vector retrieval without maintaining a separate search service.
Khelan Modi introduces the Azure DocumentDB MCP Toolkit in public preview, giving AI agents access to MongoDB-compatible database operations. The article covers client setup, read-only defaults, Microsoft Entra ID authentication, and explicit write controls, with examples for query troubleshooting, retrieval-augmented generation, and operational triage.
End of content