resume

Master's in Data Science

Harvard John A. Paulson School of Engineering and Applied Sciences

• GPA 4.0/4.0

Honors Bachelor of Science (HBSc)

University of Toronto, Toronto, Canada

• Data Science Major & Mathematics Minor
  • GPA 3.99/4.0
  • Dean's List Scholar (2020-2023)

Research Intern - Supervised by Prof. Rahul Krishnan

Vector Institute, Toronto, Canada

• Developed binary classification models using XGBoost to predict 14-day in-hospital mortality from the GEMINI dataset with 2.2 billion+ clinical data points, assessing model performance across patient groups defined by Social Determinants of Health.
• Analyzed how distribution shifts affected model performance, observing up to a 3.5% AUC improvement when deploying models trained on more socially-diverse subgroups to less socially-diverse subgroups.
• Enhanced model interpretability through SHAP analysis, identifying key features influencing predictions across diverse patient subgroups and demonstrating the need for diverse training data to improve AI model generalizability.

Research Intern - Supervised by Prof. Fanny Chevalier

University of Toronto Dynamic Graphics Project Lab, Toronto, Canada

• Developed and deployed an augmented reality application on Microsoft HoloLens to assist presenters during Q&A sessions by analyzing audience audio input and presentation context. The system generates real-time answers and relevant content to help presenters respond more effectively on the spot.
• Fine-tuned a BigBird-based question-answering model and DistilRoBERTa-based sentence transformer on the Huggingface Adversarial QA Dataset, achieving a 27% improvement in Exact Match score over the baseline model.
• Streamlined the pipeline from speech recognition and presentation context processing to answer generation and relevant content extraction. Integrated the outputs using GPT-3.5, and displayed the final response on the HoloLens screen via Unity.

Technical Sales Specialist Intern

IBM (Data & AI Team), Toronto, Canada

• Performed interactive product demos and provided support for external client Hackathons to showcase IBM’s AI solutions, leveraging a comprehensive understanding of MLOps, trustworthy AI, and cloud integration.
• Winner of the Intern Challenge - an internal client presentation competition amongst interns.

Business Intelligence Research Analyst Intern

Enverus, Calgary, Canada

• Analyzed energy operator portfolios using historical data to predict future performance, building regression models and conducting time series analysis in R to optimize strategic decision-making.
• Automated the daily model audit process using Python, reducing manual effort by 30% and improving operational efficiency.
• Employed Tableau to visualize large-scale data, transforming raw data into actionable insights and creating comprehensive reports that effectively communicated trends and strategic implications to stakeholders.

Twitter Sentiment Analysis Under COVID-19 (Python)

• Built a Naive Bayes classifier to predict the popularity of political tweets, incorporating retweets, favorites, and follower counts, achieving a 48% accuracy improvement over the baseline.
• Conducted EDA on 10,000+ tweets, engineering a custom "Popularity Score" to normalize engagement metrics and identify key trends, using Pandas and Matplotlib for data analysis and visualization.

Education Platform Student Answer Prediction (Python)

• Implemented an Item Response Theory Model to predict correct responses, modeling student ability against question difficulty to generate probability distributions.
• Optimized predictions by experimenting with Autoencoders, Matrix Factorization, Neural Networks, and Ensembles.

ASA DataFest(R) - "Best Insight" Winning Team of 2022

• Built multilevel linear mixed-effect models and correlation graphs to investigate how a player’s gaming experience of the educational game “Elm City Stories” relates to their improved efficacy in resisting drug.
• Evaluated the educational effectiveness of each game chapter and identified characteristics of the players who benefited more from the game for future game development.

• T. A. Reed Scholarship - Top cGPA in the Innis College, University of Toronto
• The Undergraduate Student Research Awards, NSERC

• First Place in the ASA DataFest, American Statistical Association and University of Toronto

• Later Life Learning Scholarship - Top 3% in the Innis College, University of Toronto
• University of Toronto In-Course Scholarship, University of Toronto