Andy Smithwick
Software Engineer & Researcher
About Me
I'm a Data Scientist II at Edwards Lifesciences and a UC San Diego M.S. Computer Science graduate specializing in AI and machine learning. My work spans machine learning, NLP research, web development, and data visualization, with a focus on building impactful solutions for the medical device industry.
I develop AI-driven solutions for healthcare and medical device analytics, including multimodal document extraction and human-in-the-loop validation workflows. I also enjoy teaching, contributing to open source, and researching how technology can better serve people.
My previous experience includes VR data visualization work with the Qualcomm Institute, UI development and accessibility work at Cornerstone OnDemand, full-stack development for ACM UCSD's membership portal serving 6,000+ users, and research resulting in publications at IEEE ISTAS, Computing in Cardiology, ITiCSE, and ASSETS.
I thrive in interdisciplinary teams, combining technical expertise with a collaborative mindset, and I'm always eager to tackle new challenges in emerging technologies.
Experience
Data Scientist II · Edwards Lifesciences
June 2026 – Present
- Deployed a TypeScript website on AWS EC2 and S3 with AWS Bedrock, using LLMs and multimodal AI models to extract key information from approximately 5,000 documents per year.
Data Science / NLP Intern & Contractor · Edwards Lifesciences
Previous experience
- Developed a Python Streamlit portal for human validation of AI pipeline output and label reuse.
- Trained Random Forest models for feature ranking and contributed to XGBoost experiments on ECG-derived features.
- Fine-tuned masked autoencoder vision transformers (ViT) for disease classification.
Education
University of California San Diego
M.S. in Computer Science, AI & Machine Learning Specialization · June 2026
GPA: 3.85 · B.S. in Computer Science: 2025
Featured Projects
Medical Device Data Analytics Platform
Developing AI and data science solutions at Edwards Lifesciences for analyzing medical device performance data and patient outcomes, including multimodal document extraction and predictive modeling.
Current RoleClinical Text Mining System
Building human-in-the-loop validation workflows and machine learning pipelines for clinical and ECG-derived data, supporting evidence-based decision making in cardiovascular medicine.
Healthcare AIVR Space Weather Data Visualization
Previously developed immersive Virtual Reality visualization of voxel space weather data using Unreal Engine 5 for the Qualcomm Institute. Created interactive 3D environments to help researchers understand complex atmospheric phenomena.
Previous WorkAndy's Donuts Mobile Game
Published mobile game available on the App Store featuring engaging gameplay mechanics and polished user interface.
App StoreMetahuman Avatars with Fine-tuned Llama3
Advanced AI avatar system combining visual metahuman technology with a fine-tuned Llama3 language model for natural conversation and interaction.
Research ProjectPrivacySyllabus-PrivacyCourseExtraction
Extracts privacy-related courses from university catalogs using Python scripts and NLP techniques.
GitHub RepoOpen Source & Community Contributions
Hackathon Resources Page Implementation
Added a comprehensive resources page to support hackathon participants with curated links and documentation.
ACM UCSD Communities Page
Contributed to the ACM UCSD Communities page, connecting students to technical, professional, and social groups across campus.
Membership Portal UI v2
Contributed to the Next.js rewrite of ACM UCSD's membership portal, improving user experience and accessibility for members.
More Contributions
Research & Publications
"Parent seeking Roblox Safety Help": Comparing Parental Roblox Concerns to Roblox Offerings
A. Smithwick, et al.
2024 IEEE International Symposium on Technology and Society (ISTAS)
Finetuning Foundational ECG Models to Detect Chagas Disease
Kelvin Nguyen, Andy Smithwick, Maxwell Loetscher, Shadi Manafi, Zaniar Adarlan, Saman Parvaneh
2025 Computing in Cardiology (CinC)
Using Peer Code Reviews to Scale a Brownfield Software Engineering Course
Third author
2026 ACM Conference on Innovation and Technology in Computer Science Education (ITiCSE '26)
Everyday Technology Practices of Adults with ADHD: Insights for Assistive Technology Design
Second author
2026 ACM Symposium on Accessible Computing (ASSETS '26)
Leadership, Teaching & Community
CSE Teaching Assistant · UC San Diego
April 2023 – March 2026
- Developed assignments in the CPython open-source codebase focused on Copilot, testing, and code review.
- Automated attendance grading to Gradescope, reducing grading time by 75%, with a 96% approval rating across six quarters.
IEEE Webmaster
November 2024 – December 2025
- Built a self-hosted CMS site with Coolify, AstroJS, and Tailwind CSS to enable non-developer content updates.
Residential Assistant · UC San Diego
- Selected as 1 of 15 from more than 1,000 applicants.
- Managed a $10,500 budget across 30+ community events.