Gary Wang
Statistics & Economics student at UC Berkeley · Business Analyst & Machine Learning Researcher
About
I'm a Statistics and Economics student at UC Berkeley, minoring in Data Science. I currently work as a Business Analyst at Alibaba's Cross-Border E-commerce Service Center, and as a Machine Learning Researcher at UC Riverside and the William and Barbara Leonard Transportation Center, where I build reward-modeling and tensor decomposition tools to improve the accuracy of a university chatbot. I'm interested in applying statistics and machine learning to real-world problems in language, finance, and education, and I've published NLP research and presented it at an international conference.
Experience
Business Analyst
Alibaba Cross-Border E-commerce Service Center · Hangzhou, China
- Built evaluation frameworks to measure experimentation efficiency and performance across AI-driven e-commerce initiatives.
- Analyzed scalable AI marketplace solutions using econometric modeling and statistical methods to inform strategic decision-making.
Machine Learning Researcher
University of California, Riverside · Riverside, CA
- Applied tensor decomposition, CP decomposition, and related machine learning techniques to classify and categorize AI-generated responses from university chatbot interactions.
- Evaluated contextual accuracy across diverse student inquiry categories, analyzing response quality and identifying performance trends in information retrieval and question answering.
Machine Learning Engineer
William and Barbara Leonard Transportation Center · San Bernardino, CA
- Developed reward-modeling pipelines and evaluation frameworks for chatbot systems, measuring confidence calibration and faithfulness across LLM-generated outputs.
- Implemented automated experiments and data analysis workflows to assess model truthfulness under varying prompts and conversational contexts.
Software Engineer
MP Biomedicals, Inc. · Santa Ana, CA
- Full-stack development on features for a customer support portal supporting nucleic acid homogenizer systems.
- Collaborated with engineers and stakeholders to test, debug, and deploy application features, improving system usability and customer support workflows.
Projects
Job Salary Prediction Model for the IT Field
Data Analyst & Developer · Pomona, CA
- Designed a salary prediction system with 10+ customizable job-description inputs using multiple machine learning algorithms including KNN and linear regression, trained on 53,000+ real-world records.
- Performed feature engineering, data preprocessing, and model validation to improve prediction accuracy and assess model performance across diverse job profiles.
Language Learning Model Research Paper
Author · Irvine, CA
- Developed an interactive, video-based language-learning app and authored a ten-page paper analyzing its design and research findings.
- Selected as 1 of 3 student speakers and 1 of 9 student authors at the ADNLP 2024 International Conference; paper later published in a special issue across 6+ NLP journals.
This platform
goat-ai-platform · gariyuuu.com
- Self-hosted, OpenAI-compatible AI gateway (auth, rate limiting, usage tracking, LoRA fine-tuning pipeline) fronting an open-weight model, replacing direct calls to proprietary LLM APIs across a dozen personal apps.
Education
University of California, Berkeley
B.A. Statistics & Economics, Minor in Data Science
- Relevant coursework: STAT 134 (Concepts of Probability), STAT 133 (Concepts in Computing with Data), MATH 54 (Linear Algebra & Differential Equations), MATH 53 (Multivariable Calculus), ECON 140 (Econometrics), OIDD 0001 (Prescriptive Analytics: Business Optimization, UPenn)
Skills
Honors
- ADNLP 2024 recognition of Language Learning App research paper and lead speaker
- US Patent on Foreign Language Learning App and NLP methods
- First Place, Ignite CS Programming Expo
Contact
The best way to reach me is by email at gywng006@gmail.com, or connect with me on LinkedIn. You can also check out my code on GitHub or my published research on OpenReview.