• Languages: C++ (Low-latency/Modern C++), Python (Data Science), SQL(ClickHouse)
• Libraries/Frameworks: Eigen, OpenMP, STL, NumPy, Pandas, light-gbm, PyTorch,langchain, langgraph
• Infrastructure & Tools: Linux, Git, Docker
• Math & Finance: Stochastic Analysis, Time Series Analysis, Convex Optimization, Market Microstructure

[Beijing Key Laboratory of Financial Artificial Intelligence] | Core Researcher | Apr 2025 – Mar 2026

• Built multi-agent LLM systems for banking operations, integrating internal data warehouses and external platforms; designed regulatory knowledge graphs, tool APIs, and fine-tuning QA datasets; fixing vulnerabilities in dependencies detected by Fortify scan.

• Developed CUFEL-A, a multi-agent framework for automated A-share research report generation with a paper-planning–architecture-report structure , web search tools, and prompt optimization.

• Built CUFEL-Q Arena, a multi-agent quantitative trading platform with multi-frequency financial data pipelines, NLP/TDA text indicators, and high-precision backtesting engines.

• Developed physics-informed deep learning models for Tokamak nuclear fusion, combining TimeXer encoders, CNN/FNN decoders, and autoregressive forecasting.

[Minghuhui Private Investment Fund] | High-Frequency Quantitative Research Intern | Aug 2024 – Jan 2025

• Built vectorized high-frequency backtesting engines in Python; accelerated intraday simulation via matrix operations and Dask parallelization.

• Constructed 100+ LOB and trade-flow factors using Python + ClickHouse; evaluated predictive power via IC/ICIR.

• Developed DeepLOB and LightGBM trading models for convertible bonds and BTCUSDT T0 strategies.

[Huatai Securities] | Quantitative Research Intern | Aug 2023 – Apr 2024

• Developed a Python backtesting framework for equity strategies with portfolio construction and IC/RankIC evaluation.

• Implemented analyst forecast surprise strategies for A-share equities and conducted empirical backtests.

• Built tools for exchange-traded option strategy analysis, including implied volatility estimation and Greeks evaluation.

• Description: Factor Research Series 1: Genetic Programming Factor Mining Algorithm implemented in C++.
• Tech: C++20, OpenMP, Eigen

• Description: Factor Research Series 2: LLM-Driven Multi-Agent System for Explainable Alpha Discovery.
• Tech: langchain, function call

• Description: A General Backtest Framework for Quantative Research.
• Tech: python, numpy, pandas

• Description: A ETF Strategy based on PTree data selection for stable performances.
• Tech: PTree, iTransformer

• Description: This project explores the time-series predictability of the HS300 ETF by leveraging a "zoo" of 105 high-frequency characteristic-sorted portfolios.
• Tech: jump-diffusion decomposition, high-frequency trading

• 🔬 [QuantMoE Model]: Mixture of Experts DL Models for stocks trading with a specified optimization function.
• 📚 [Market Regimes Detector]: Using Reinforcement Learning and Markov probability transition matrix to predict Market Regimes.
• 🧠 [Hybrid Data Factor Miner]: Using multi-frequency trading data and nlp indicator to discover new factors.