Experiences

Experience

Selected engineering, research, and applied AI work from my resume.

Terasaki Institute for Biomedical Innovation

Jr Biomedical Intern

March 2025 - Present Los Angeles, CA
PyTorch XGBoost ClinicalBERT Epic CDS
  • Built a kidney transplant rejection prediction model for integration into Mayo Clinic's Epic clinical decision support system.
  • Implemented PyTorch model with XGBoost and neural network as backbone, and variational autoencoder for oversampling.
  • Experimented with the use of OpenAI's gpt-oss and ClinicalBERT in classifying 2 kidney allograft rejections (ABMR and TCMR).
  • Collaborated with 6 team members including machine learning scientists, transplant experts, physicians, and Mayo Clinic's AI team.

University of Virginia

Machine Learning Researcher

April 2022 - Present Charlottesville, VA
Time Series LSTM Hugging Face IR
  • Advisor: Geoffrey Fox (Professor of Computer Science); Judy Fox (Associate Professor of Data Science).
  • Standardized a data-driven approach to spatio-temporal series analysis demonstrated by the use of LSTM in rainfall-runoff modeling.
  • Designed and led the implementation of an agentic model recommendation system for the Time Series Commons project.
  • Benchmarked 25 time series foundation models on 27 datasets from Hugging Face to scaffold the knowledge base for the agent.
  • Prepared JSON schemas and indexing registries for time series models, datasets, and benchmark runs for information retrieval.

DealCompass

Software Engineer Intern

June 2024 - July 2024 South Bend, IN
React REST APIs Slack Gmail
  • Integrated Slack and Gmail features into the real-estate investment CRM software with the React framework and REST APIs.
  • Developed the front-end to mimic actual Slack and Gmail software utilizing JavaScript, HTML, and CSS.

Zhongtiao Technology Co., Ltd.

AI Engineer Intern

December 2023 - January 2024 Shanghai, China
Qwen-14B RLHF Redis OCR
  • Performed prompt engineering and reinforcement learning with human feedback on Alibaba Cloud's Qwen-14B foundation model.
  • Collaborated with a 10-person Agile team to fine-tune a customer service chatbot for Geely Auto, Cadillac, and Volkswagen.
  • Built an automatic chatbot feedback testing algorithm utilizing optical character recognition, the Redis database, and Qwen API.