Tag
This paper presents a deployed pipeline for semantic user profiling that uses per-transaction-pattern LLM inference to reduce costs by three orders of magnitude at a major Japanese bank.
ProfiLLM introduces an agentic LLM pipeline that generates utility-aligned user profiles from platform-scale behavioral logs for industrial ride-hailing dispatch, achieving significant improvements in outcome prediction and GMV in production at DiDi.
PaperFlow is a framework for scientific paper recommendation that processes user profiles, daily paper streams, and interest drift through three stages: profiling, recommending, and adapting, evaluated on a longitudinal benchmark with 24 users and 50 daily streams.