Build a Recommendation System and Conversational Interface from Scratch

Starting from a blank slate, build the first version of a recommender engine. After that core system is defined, add a conversational chat layer that can retrieve suggested items and explain them. Every assistant response must respect the same item-eligibility and ranking rules as the underlying recommender. Assumptions: There is no prior recommendation infrastructure already in place. The assistant may only reference items that actually exist in the catalog, and must never…