Vector Indices
Vector databases power AI applications through similarity search. They are a key component of many AI applications, including recommender systems, search engines, and chatbots. You can store embeddings in a vector database and then query it to find the most similar embeddings to a given query embedding.
ModelFusion provides a VectorIndex interface that can be implemented by different vector databases. This allows you to use the same code to query different vector databases, e.g. Pinecone or an in-memory vector store. Learn more...