CV

Research Assistant (Advisor: Prof. Xiaowei Luo)

City University of Hong Kong

Climate Risk Assessment System.

• Applied the ISI-MIP ATTRICI framework (Python, xarray, netCDF4, PyMC/SciPy) to generate counterfactual climate datasets for aviation emission trade-off risk assessment.
• Ran the 8-step detrending and quantile-mapping pipeline to fit variable-specific statistical models and preserve observational rank structure across scenarios.
• Built probabilistic risk-analysis and visualization workflows to assess CO2 vs. non-CO2 aviation emission trade-offs for research decision support.

Research Assistant (Advisor: Prof. Mi Xiang)

Shanghai Jiao Tong University

Aviation Personnel Mental Health Tracking Application.

• Implemented a session-memory component for a mental-health support agent using LangGraph, FastAPI, and PostgreSQL/pgvector with CoALA-inspired memory design.
• Structured memory into three CoALA-inspired layers—semantic facts, episodic arcs, and procedural rules—for longitudinal mental-health tracking.
• Connected memory and safety logic across five interfaces (CLI, Next.js web chat, LiveKit web voice, Telegram, FastAPI backend).

AI Software Developer Intern

Zhongshan Securities Co., Ltd

• Built a heterogeneous fraud-graph preprocessing pipeline (Python, pandas, DGL, SageMaker) over the IEEE-CIS fraud dataset for transaction-risk classification.
• Trained graph-based fraud classifiers with DGL, comparing GNN relational modeling against MLP and XGBoost baselines.
• Configured an end-to-end AWS deployment workflow with Lambda triggers, custom Docker images, S3, IAM, and CloudFormation.

Fullstack Intern

PhinD LLC

• Adapted AWS’s sagemaker-graph-fraud-detection solution, running its preprocessing pipeline on IEEE-CIS data for fraud classification.
• Evaluated graph-based fraud classifiers using DGL heterogeneous graphs and RGCN against MLP and XGBoost baselines.
• Deployed the end-to-end AWS workflow with Lambda-triggered modules, Docker images, S3, IAM, and CloudFormation.

AI Software Developer Intern

NetEase Youdao Information Technology Co., Ltd.

• Built a multilingual OCR engine supporting 20 languages at 96% per-image accuracy, 1.5s/image, 100K+ daily volume.
• Developed OpenCV-based image enhancement and YOLOv8 text detection (94.2% localization), improving blurred-image recognition by 30%.
• Built a multi-engine translation service (FastAPI, Google/Baidu/DeepL) supporting 50+ languages at 300ms average latency.
• Developed an AI text-polishing system on GPT-3.5 with prompt engineering, improving readability by 30% and reaching 89% user satisfaction.

B.S.

University of Illinois Urbana-Champaign

Statistics (Minor: Mathematics)

Undergraduate Research Symposium (Submitted)

Beixuan Jin, Dylan Wu, et al. University of Illinois Urbana-Champaign.