This is a comparison of Approximate Nearest Neighbor (ANN) speeds between Pinecone and several flavors of Elasticsearch. Thanks to Dmitry Kan, whose article is referenced here, for reviewing drafts of this post and providing valuable suggestions.
Semantic search applications have two core ingredients: Text documents represented as vector embeddings that capture meaning, and a search algorithm to retrieve semantically similar items.
This is a companion discussion topic for the original entry at https://www.pinecone.io/learn/bert-search-speed/