We’ll definitely have this at some point. A good recommendation system needs a solid base of user data, so it’s only possible once we have a larger user base. I really wanted to have this for the launch since it seemed so important for a book tracking platform but we didn't have the data to train models to make recommendations.

I mean, not really you don't need user data; I was doing something related with LLM (simply asking it) and planning to use vector databases for various tasks, like semantic search, similarity matching, clustering, and topic modeling. Later pivoted to doing something for podcasts (analysing transcripts).