CV

Youxing LI

rickliyouxing@163.com
Hong Kong / Shenzhen, , CN

Summary

Researching Multimodal Learning, Riemannian Deep Learning, and Data Mining.

Education

  • Exchange Program in Quantitative Finance
    2023-06-30
    Hong Kong University of Science and Technology (HKUST)
  • BSc in Quantitative Finance — Jinhe Center for Economic Research
    2025-06-30
    Xi'an Jiaotong University (XJTU)
    GPA: 87.04/100

Work Experience

  • Agent Algorithms Engineer
    2025-09-01 - 2026-02-28
    Bauhinia AI Limited / Prof. Kani Chen's Lab @ HKUST
    Led end-to-end development of TradingNews.press, an AI-powered financial news analysis platform with multi-agent workflow orchestration.
    • Led end-to-end development of TradingNews.press from inception, encompassing algorithm design, system architecture, and production deployment.
    • Architected CRDR workflow orchestration (Channel→Relevant→Decision→ReAct) using LangGraph state machines with parallel processing.
    • Developed 7 specialized financial analysis agents with custom prompt engineering and Pydantic output schemas.
    • Implemented text-grad architecture with stock price tracking and prompt propagation computation graph via LLM-as-a-judge framework.

Skills

Research Areas

  • Multimodal Learning
  • Reinforcement Learning
  • Data Mining
  • Riemannian Deep Learning
  • Optimal Transport

Technical

  • Python
  • LangGraph
  • LLM Agent Development
  • Pydantic
  • Prompt Engineering
  • Riemannian Geometry

Publications

  • Riemannian Sliced Wasserstein Distance on Correlation Matrix
    2026
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026)
    Youxing, L., & Zihen, C. (2026). submitted to IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026). Review scores: 5/4/5. Confidence: 4/2/2.
  • Geometry-Aware Second-Order Pooling for Fine-Grained Visual Recognition
    2026
    Under review
    Youxing, L. & Zihen, C. currently under review.

Portfolio

  • TradingNews.press — AI Financial News Platform
    2025
    Portfolio
    AI-powered financial news analysis and investment signal generation platform built with multi-agent workflow orchestration.
  • Tsinghua Economic Hotspot Analysis Competition — Top 1%
    2023
    Portfolio
    Led team research on AI-enabled coal mining transformation. Built Cobb-Douglas model, conducted field research at 6 enterprises.

Languages

  • Chinese
    Native
  • English
    Professional

Interests

  • Hobbies
    Music, Playing the Drum Kit, Fitness, Running