Databases centered around use cases related to word embeddings, usually for near-text search (see Text Indexes for more classic techniques) or similar use-cases.
I've used this in the past. It might be tricky getting used to some of their quicks, but I don't have any specific reference points to compare it with.