Resume

Download PDF Version

Education

University of British Columbia (UBC) | Vancouver, BC
M.S. in Data Science, Computational Linguistics | Aug 2025 - Present

  • Relevant Coursework: Supervised Learning, Unsupervised Learning, Natural Language Processing, Machine Learning, Statistics.
  • Academic Focus: Examining data patterns and building models to predict the future.

University of California, Santa Cruz (UCSC) | Santa Cruz, CA
B.A. in Business Management Economics | Sep 2019 - June 2022

  • Relevant Coursework: Advanced Quantitative Analysis, Machine Learning Economics, Statistics, Security Market.
  • Academic Focus: Analyzing economic patterns using Machine Learning and Statistics.

Technical Skills

Category Skills
Languages Python, R, SQL, MongoDB, Bash
ML & AI PyTorch, Scikit-learn, XGBoost, Hugging Face Transformers, BERT, RoBERTa, Flan-T5, sentence-transformers, LLM Fine-tuning, RAG, Multi-agent LLM, Vector Embeddings, Prompt Engineering, SHAP
NLP Word Embeddings, N-gram Models, Parsing Algorithms, Membership Inference, NLTK, spaCy, CEFR
Data Engineering Pandas, NumPy, Apache Parquet, ETL Pipelines, Feature Engineering, Data Visualization
MLOps & Cloud Docker, AWS, Git / GitHub, CI/CD, Weights & Biases (W&B), LangChain / LlamaIndex
Tools VS Code, Jupyter, Quarto, LaTeX

Experience

Applied Machine Learning Engineer Intern | Ednoda x UBC MDS-CL Capstone | Jan 2026 - June 2026

Built an end-to-end AI-powered curriculum content pipeline for an ESL edtech platform, spanning grammar analysis, semantic retrieval, and automated teaching material generation.

  • Designed and implemented 29 spaCy DependencyMatcher grammar templates to classify ESL sentences by syntactic structure, achieving 65.8% coverage on production sentence data and 54.2% on the CEFR-SP benchmark corpus (A1–C2 consistent)
  • Built an offline embedding pipeline using all-MiniLM-L6-v2 to encode 6,477 grammar-tagged sentences into a searchable semantic index
  • Fine-tuned a BERT model for CEFR difficulty classification (A1–C2), achieving 92% (±1) accuracy
  • Engineered a three-layer sentence retrieval system combining grammar structure filtering, CEFR difficulty matching, and cosine similarity ranking to surface curriculum-aligned example sentences
  • Built a three-layer content generation system that produces new ESL practice sentences by detecting replaceable slots, generating masked-token candidates (RoBERTa for single-token slots, T5/Flan-T5 for multi-token spans), and ranking candidates by cosine similarity for fluency and difficulty fit
  • Engineered a PDF parsing pipeline to extract and structure question sections from textbook PDFs into machine-readable question datasets
  • Built a RAG (Retrieval-Augmented Generation) system that retrieves semantically relevant existing questions to reduce redundancy and improve consistency across generated content
  • Designed a multi-agent LLM workflow that ingests a textbook PDF and automatically produces a complete teaching package — structured lesson notebook, practice questions, and graded quiz — ready for classroom use

Career Mentorship Program | Mentee | Jan 2026 - Present

  • Selected for a professional mentorship program under industry veteran Jacob Oh.
  • Focusing on software engineering best practices, career development in tech, and bridging academic concepts to industry standards.

Capstone

Ednoda x UBC MDS-CL Capstone — Applied Machine Learning Engineer Intern

Built an end-to-end AI-powered curriculum content pipeline for an ESL edtech platform, spanning grammar analysis, semantic retrieval, and automated teaching material generation.

Tech Stack: Python · spaCy · sentence-transformers · BERT · RoBERTa · Flan-T5 · LangChain / LlamaIndex · RAG · Multi-agent LLM · pandas · Apache Parquet · NLP · CEFR