Shane Williams

Masters student in Bioinformatics with 2+ years experience in Fullstack Software Development, interested in pivoting to a Machine Learning Engineer role in Healthcare. Proficient in TensorFlow and Python, with academic experience building models from medical-imaging datasets. Hands-on experience deploying and scaling deep learning models on cloud platforms like GCP and AWS.

Tufts Machine Learning Research Lab | Machine Learning Researcher | Sept 2024 - Current

• Researched semi-supervised and transfer learning approaches to predict aortic stenosis severity from low-quality Point-of-Care Ultrasounds (POCUS) under the supervision of Dr. Michael Hughes.
• Implemented Gaussian Process Abstention in our Multi-Instance Learning model to enhance distance-awareness, allowing the model to better handle out-of-distribution data, giving clinicians greater confidence in its performance on real-world datasets.
• Developed additional performance metrics and analytical insights to support the integration of SAMIL into clinical practice, focusing on understanding the potential impact on healthcare resources.

Lute.ly | Full Stack Engineer | Sept 2023 - Dec 2024

• Integrated Stripe Direct Payments to enable customer-to-customer purchases and subscription management for recurring payments on the Lute.ly web platform.
• Designed and implemented label account features, utilizing AWS Cognito and JWT token authorization to enable secure parent account functionality.
• Enhanced the performance of backend genre classifiers and led data curation initiatives to improve the accuracy of convolutional neural networks (CNNs).

Major League Baseball | Software Engineering Intern | May 2023 - Aug 2023

• Created NuxtJS/VueJS web application for all Mobile Services API’s and documented all team repositories with Swagger.
• Increased code coverage for Mobile Team Github organization on average by 30.1%, became proficient using Golang Testify and Mockery for advanced end-to-end API testing.
• Implemented Google Cloud Function python script using JIRA REST API and GCP API Services to automate sprint analysis saving conservatively 48 hours of manual labor per Project Manager.

Tufts University | Masters of BioEngineering: Bioinformatics | Sept '24 - May '26

Johns Hopkins University | Bachelors of Science in Computer Science | Sept ‘18 - May ’22