AWS ECS engineer working on cloud infrastructure.

At Sohn Lab, I am developing methods using LLMs and statistical inference to improve patient engagement and outcomes when engaging with radiological reports and results.

Built a perceptual-hash-based image comparison service in AWS cloud and integrated the service with existing reliability monitoring infrastructure to create a broader view of organization-wide system health metrics and reliability of multi-media delivery processes. Implemented the service with a combination of AWS tools and services such as Lambdas, Queues, and EC2 instances by using a declarative infrastructure specification framework in Typescript to establish a CI/CD pipeline and then using Java along with the unit testing framework JUnit and the direct-injection framework Dagger.Explored variations of image compression and corruption as well as alternative image similarity comparison methods by SVD/PCA, classic hand-crafted edge detectors, support vector machines, and neural image embeddings and compared their relative effectiveness and computational cost. Used Python, Numpy, OpenCV, Scikit-Learn, and PyTorch in a Jupyter Notebook to carry out investigations.

Worked on Vue.js landing pages, backend service using Laravel and PyroCMS, and database query design using MySQL.