cv简历

Master of Computational Data Science (MCDS)

Carnegie Mellon University

Analytics Track

Bachelor of Science, cum laude

New York University Shanghai

Double Major: Computer Science, Mathematics

• Cumulative GPA: 3.793/4.0 | CS Major GPA: 3.946/4.0 | Math Major GPA: 3.786/4.0
• Relevant Coursework: Honors Linear Algebra, Basic Algorithms, Operating Systems, Databases, Natural Language Processing

Generative Retrieval for Multimodal Recommendation System

Research Assistant, New York University Shanghai, Shanghai, China

• Advisor: Prof. Hongyi Wen
• Developed a multimodal framework for sequential recommendation on Amazon datasets. Identified and applied VQ-VAE, RQ-VAE, OPQ to capture discrete behavior patterns, achieving 47.2% improvement in NDCG@10 (0.0339 to 0.0499).
• Collaborated with 2 teammates and implemented BPE-inspired partitioning tokenization across item sequences, merging frequent sub-sequences to enhance contextual coherence and representation quality.

Energy-Aware Route Optimization Under Traffic-Light Timing

Research Assistant, New York University Shanghai, Shanghai, China

• Advisor: Prof. Zhibin Chen
• Proposed and designed a traffic-aware routing system under real-world map constraints, minimizing energy consumption.
• Built an OSM, SUMO, TraCI pipeline and extended Dijkstra's algorithm to multi-objective (time, energy) shortest paths, integrating electric and gasoline vehicle energy models. Achieved 10–15% fuel savings at 2–5% time cost.

Efficient Retrosynthesis with Pretrained LM Fine-Tuning

Center for Data Science, New York University

• Trained predictive models for chemical reaction retrosynthesis using SMILES notation, applying an LLM (LlaSMol-Llama2) for dataset reconstruction and augmentation, improving Top-1 Accuracy by 26.2% (0.588 to 0.742).
• Explored and employed multiple fine-tuning strategies (encoder freezing, MLP adapters, LoRA) to identify optimal approach with best accuracy-efficiency trade-off.

Fantasy Sports League Database

Tandon School of Engineering, New York University

• Self-directed a project to design a 13-table MySQL database populated with 390+ records, supporting multi-league sports with user authentication, draft scheduling, roster management, and trading systems. Engineered a Python backend and HTML/CSS frontend with ACID-compliant transactions, triggers, and stored procedures, integrating full-stack functionality.

Data Engineer Intern

Data Strategy Department, McDonald's China

Worked on building and optimizing data infrastructure to support business intelligence and analytics for McDonald's China.

• Spearheaded an internal AI agent for self-service data access, routing natural-language queries to constraint-aware SQL over a catalog spanning 14 business domains, or RAG over internal data maps and documents to generate cited answers.
• Initiated and designed a similarity-based sales forecasting algorithm over a combo master-data set normalized across 4 ID levels, quantifying menu-composition to forecast new-combo launch sales from comparable combos.
• Managed and validated enterprise database schemas, governing a 203-scenario and 54-interface data map.
• Redesigned data dashboards, organizing 500+ financial metrics with structured drill-down paths for self-service analytics.

Treasurer

PEER-Pal Club, New York University Shanghai

• Managed financial operations and budgeting for a club of 100+ club members, mentored incoming club leaders, ensuring continuity of club's charity work; facilitated letter exchanges with 60+ high school students from Hunan Province.

Trombone Section Leader

Chamber Orchestra, New York University Shanghai

• Led and trained trombone section, coordinating weekly rehearsals and semester concerts, and oversaw instrument logistics.

Founders' Day Award

New York University Shanghai

Distinguished Graduate Award.

Dean's List

New York University Shanghai

Academic years 2022–2023, 2023–2024, 2024–2025 (GPA ≥ 3.7).

Mathematical Contest in Modeling, Certificate of Achievement, Honorable Mention

COMAP

Preprocessed large datasets to remove anomalies and ensure data quality; built predictive models using methods including Bayesian regression for regional market data analysis.