aurorazeng

• Led the development of an Agentic RAG system for automated financial literature analysis, reducing average reading time from 45 minutes to 10 minutes per paper. • Designed and implemented a distributed web crawler using Selenium to collect and maintain an incremental database of top-tier financial journals (JF, JFE, RFS). • Architected a multi-modal data infrastructure integrating MySQL, MinIO, Redis, and FAISS to enable scalable storage and semantic retrieval of 26,000+ research papers. • Integrated OpenAI API with optimized prompt engineering to generate structured summaries, key insights, and academic highlight extraction with 90% accuracy. • Built RESTful APIs with FastAPI and developed a responsive Vue3 front-end supporting PDF split-screen analysis and citation network visualization. • Served 1,700+ active users and improved literature screening efficiency by 70%.