Designer | Developer | Creator
I am a graduate student in Operation Research at Columbia Engineering. My research area focuses on Machine Learning and Financial Engineering. Prior to joining Columbia Univerisity, I got my Bachelor's degree in Computer Science at the University of British Columbia.
I have experience in data science, quantitative finance, machine learning, and software development through various internships. Currently, I am working as a AI Engineer Intern at Retell AI, focusing on AI generation and building LLM-powered AI voice agents.
You can find my up-to-date CV here.
· Core courses: Probability and Statistics, Machine Learning, Stochastic Process, Simulation, Optimization Models, Quantitative Finance
New York, USA
Sept. 1, 2023 - Dec. 31, 2024
· Cumulative GPA: 3.8, Dean’s List Honors · Core courses: Software Engineering, Data structures and Algorithms, Linear Algebra, Calculus, Web Development, Operating Systems, Data Mining, Machine Learning, Microeconomics, Macroeconomics, Introduction to Artificial Intelligence
Vancouver, Canada
Sept. 1, 2019 - May 1, 2023
· Developed over 10 customizable, LLM-based models for AI voice agents, seamlessly integrating Generative AI APIs such as Claude and GPT-4o to enable natural language understanding and real-time, conversational response generation.
· Explored and developed an automatic speech recognition system for speech-to-text transcription, achieving a real-time processing latency of under 50 milliseconds. Integrated and fine-tuned a pre-trained deep learning model for speech emotion detection.
· Conducted comprehensive internal A/B testing on platform features, analyzing over 1,000 simulated user interactions.
· Developed RESTful APIs to support seamless integration with web applications. Built 10+ interactive web interfaces using React and Next.js, featuring dynamic real-time data visualizations for user usage that drove a 20% increase in engagement.
Redwood, CA
Sept. 1, 2024 - Dec. 20, 2024
· Assist in developing teaching materials for the graduate-level course Optimization Models and Methods in Financial Engineering, covering topics such as linear programming, stochastic optimization, and dynamic programming
· Conduct office hours and lead comprehensive review sessions prior to exams to support student learning
New York, NY
Sept. 1, 2024 - None
· Developed an AI-driven subtitle generation platform, with NeuralSpace and Whisper API for seamless end-to-end functionality.
· Designed an automated testing pipeline in Python to measure processing time and latency across multiple languages, optimizing performance through API parameter tuning and audio preprocessing to achieve nearly 100% accuracy across 100+ dialects.
· Designed a PostgreSQL database on Amazon RDS for structured user data management. Analyzed word frequency to identify domain-specific high-frequency terms, achieving a 15% increase in model accuracy.
· Developed an advanced search functionality by integrating transcription-based text search with frame-level object detection using CLIP neural networks in PyTorch, enabling users to retrieve videos based on specific keywords or visual content.
New York, NY
June 3, 2024 - Sept. 1, 2024
· Designed and implemented 5 ETL pipelines to automate the daily data extraction and cleaning process of Tableau dashboards, reducing data processing time by 30%. Engineered 10+ advanced features using Python to enhance analytical insights.
· Implemented a GRU-based model using TensorFlow for time series analysis to accurately predict future trends in
macroeconomic indices, achieving an RMSE of 0.15. Leveraged SHAP to interpret feature influence on predictions.
· Developed and backtested market-neutral arbitrage strategies to exploit stock-level pricing inefficiencies, leveraging statistical arbitrage and multi-factor models using Python. Improved strategy Sharpe ratios by 20% on average.
· Conducted comprehensive hyperparameter tuning to optimize the baseline model, achieving a 6% improvement in accuracy.
Shanghai, China
July 1, 2023 - Nov. 1, 2023
· Designed and implemented multi-factor financial models in Python to predict stock returns, including adapting the Barra China Equity Model. Utilized factors such as momentum, value, and volatility, achieving a 12% improvement in predictive accuracy.
· Extracted and refinedfinancial indicators from large-scale datasets using SQL and integrated workflows for data processing, leading to a 15% improvement in model prediction accuracy and enhancing the profitability of trading strategies by 8%.
· Identified and resolved technical issues in algorithmic trading models implemented in C++ and Python, ensuring seamless execution in live trading environments and optimizing code efficiency and reducing execution latency by 15%.
Beijing, China
May 1, 2022 - Sept. 1, 2022
· Developed an interactive PowerBI dashboard with advanced filters and drill-down functionalities, enabling detailed monitoring and analysis of video statistics to support data-driven decisions, and enhance content strategy effectiveness.
· Built an XGBoost model to analyze user characteristics and predict game registration rate and purchase likelihood. Conducted user behaviour analysis to optimize ad spend and traffic distribution, achieving a 50% increase in the new registration rate.
· Designed and executed 5+ A/B testing experiments to evaluate and compare the effectiveness of various operational strategies.
Beijing, China
April 1, 2021 - July 1, 2021
· A Data-Driven Perspective of Financial Markets.
· Detect market regimes based on 15 daily macro-signals since 1970 with Hidden- Semi Markov Model and unsupervised machine learning techniques
· Perform the topological networks with data visualization
Techniques: Python
New York, NY
Jan. 20, 2024 - June 1, 2024
· Predicted the state (up, down, stationary static) of the midprice of stocks with a high-frequency Limited Order Book and several classification models, such as the Random Forest Classifier.
· Collaborated with Mathworks
Techniques: Matlab
New York, NY
Sept. 1, 2023 - Dec. 1, 2023
· Developed several supervised regression models, such as the Linear Regression and SVM.
· Predicted the default payments of 1000+ credit card holders in Taiwan, to reduce over-issued credit and optimize the bank operation
Techniques: Python
Vancouver, Canada
Jan. 1, 2023 - May 1, 2023
· Created Simple Random Sampling and Stratified Random Sampling with R programming to estimate the average prices of Airbnb listings in NYC
· Compared sampling methods with factors, such as standard error, finite population correction and CI
Techniques: R
Vancouver, Canada
Sept. 1, 2022 - Dec. 1, 2022
· A carpool application in Vancouver, using MongoDB as database and Express web framework for server implementation
· Integrated Google API and Nodemailer for route display and email notification
· Deployed on heroku
Techniques: Javascript, Material-UI, Nodejs, HTML, React, Redux, Heroku
Vancouver, Canada
May 1, 2022 - Sept. 1, 2022
· Developed an information tracking system for the COVID-19 virus on the campus, including tracking 20k+ students' information, COVID-19 test results, etc
· Used SQL syntax to realize query functions, simulated a database and built visual local web pages
Techniques: SQL, Oracle, PHP, CSS HTML
Vancouver, Canada
Sept. 1, 2021 - Dec. 1, 2021
· Built a data processing and filtering system for campus information that featured functions to select and view average scores of any course, prerequisite courses, and detailed information on campus buildings
Techniques: TypeScript, JavaScript, Bootstrap, Mocha
Vancouver, Canada
Sept. 1, 2021 - Dec. 1, 2021
· An application that allows users to add, delete and change tasks. Additional functions including setting reminder before the deadline, timing, etc.
Techniques: Java
Vancouver, Canada
Jan. 1, 2020 - May 1, 2020
· Developed a predictive model for cybersecurity risk analysis, using ARIMA model to forecast loss frequency (incident occurrence over time) and Monte Carlo Simulation to estimate loss magnitude ($ impact.
Techniques: Python
New York, NY
Sept. 20, 2024 - Oct. 19, 2024
· Launched a special meal plan to address food insecurity problems caused by financial difficulties on campus, which efficiently solved the food insecurity problem on the UBC campus and prevented physical contact during the COVID-19 pandemic · Final ranking: 2/150
Vancouver, Canada
Jan. 1, 2021 - Jan. 31, 2021
· Developed a model trained on a large dataset of keystroke logs that have captured writing process features
New York, NY
Nov. 1, 2023 - Jan. 1, 2024
· Analyzed data with Excel Pivot Table and derived the reason for the decrease in customers in "Best Buy" offline stores. · Developed the “Saving while Walking” plan to attract customers to offline stores by releasing coupons based on customers’ in- store visit step-numbers · Final ranking: 3/200
Vancouver, Canada
Jan. 1, 2020 - Jan. 15, 2020
Intermediate
Intermediate
Intermediate
Intermediate