A self-hosted feed that fetches every new paper in Quantitative Finance (q-fin.*) and Econometrics (econ.EM) every day, runs each abstract through an LLM for a structured summary in Chinese, and exposes the result through a clean web UI.
Why — Reading raw arXiv listings every morning is friction. The summarised feed gives a 30-second triage of what is new, what matters, and what to read in full.
Stack — Python (Scrapy), DeepSeek API, vanilla HTML/JS, nginx static serve.