Breaking it into offline vs. online environments, and candidate retrieval vs. ranking steps.
A whirlwind tour of bandits, embedding+MLP, sequences, graph, and user embeddings.
An overview and comparison of the various approaches, with examples from industry search systems.
Access, serving, integrity, convenience, autopilot; use what you need.
Why real-time? How have China & US companies built them? How to design & build an MVP?
What questions do they answer? How do they compare? What open-source solutions are available?