I've created a little tool to do a semantic search of the books on Read This Twice. I'm using Open AI embeddings to create vectors from the book descriptions and user reviews, which I then store in Pinecone. Whenever a user makes a search request, it's vectorized and then used to search Pinecone.
