Microsoft has introduced a new library for .NET, called Microsoft.Extensions.VectorData.Abstractions, which aims to simplify the integration of .NET solutions with the AI Semantic Kernel SDK. This library provides abstractions over concrete AI implementations and models, making it easier for developers to build applications that utilize language modeling (LLM) technology.
The library focuses on abstracting away vector stores used for LLM embeddings, allowing developers to store and query data in a high-dimensional vector space. This enables semantic search capabilities, rather than simple text matching.
Developers can use this library to create POCO objects annotated with vector attributes, which are then stored and embedded using the IEmbeddingGenerator and IVectorStore interfaces provided by the library. The resulting embeddings can be used for querying and retrieval-augmented generation (RAG) applications.
The library is currently in preview and will remain so until .NET 9 is released. Developers can provide feedback through the GitHub repository issue list.