VectorDatabase

The encoding / embedding (LanguageEmbeddings) part of LanguageModels is pretty clever in its own right. So why not make databases out of it?

Basically takes "unstructured" data, encodes it in vectors which have some kind of similarity metric with other similar things. (The algorithms extract such similarities)

https://towardsdatascience.com/how-i-turned-my-companys-docs-into-a-searchable-database-with-openai-4f2d34bd8736 (Compare : PersonalizedChatInterfaces)

https://www.pinecone.io/

https://weaviate.io/

https://vald.vdaas.org/