Anna Arnett

This past summer, I worked on 84.51°'s AI Enablement Team, where I got hands-on experience with how companies actually deploy and monitor AI models in production! I helped make it easier for teams across the company to use machine learning by improving how we serve models with NVIDIA Triton, adding tools for logging, testing, and alerting, and building demos that made AI models more accessible for non-technical users. Along the way, I learned a ton about model serving, infrastructure, and what it really takes to operationalize AI at scale.

After my work concluded over the summer, I was selected to receive the Microelectronics Fellowship Award from the SUPREME Research Center. I am currently continuing my work with Dr. Niemier throughout the upcoming school year.

Over summer of 2024, I collaborated with amazing Notre Dame professors and PhD students to study Bayesian neural networks. BNNs provide a measure of uncertainty in their predictions, which lets us know how confident the model is in its output. I helped to develop and train models using PyTorch and TensorFlow, focusing on adding Bayesian layers to Convolutional neural networks.

I also developed a data visualization tool in Python that reads data from an Excel spreadsheet. This contributes to benchmarking efforts of the research group by facilitating easy monitoring of key performance indicators. We were able to present this tool to liaisons from Intel and the SUPREME research center, garnering further interest as well as great feedback.

I worked at NASA's Goddard Space Flight Center on the GMSEC (Goddard Mission Services Evolution Center) Team. GMSEC is a software framework that facilitates standardized messaging between space mission systems. This was an amazing first foray into the professional software engineering world! I gained experience in deploying software in Agile-based environments as well as reading/writing technical documentation.

Much of my project was focused on creating a deployment strategy using Robot Framework, Docker, and Selenium that could be expanded upon to fit any test scenario. This way, the GMSEC team could have a standardized strategy when it comes to developing test scripts for software suites of containers.

IrishSat is Notre Dame's space systems design team. We are working towards launching ND's first satellite! I am a member of the Computing Team, and we collaborate with a range of other disciplines (Engineering, Business, Physics, etc.) to fulfill all software needs for the team. This has included network programming, translating code from Python to C, and doing research on the Kalman Filter. We also have a banquet at the end of every year where we present our progress to the Notre Dame community and beyond!

CS for Good is a student-led club at Notre Dame that partners with local non-profits to provide any technical assistance they need. I have gotten to work with great organizations and I love learning more about the social good that we can do with CS! I help with matching teams to organizations, aligning club resources, and technical tasks (such as digitizing files, updating websites, and conducting data analysis).

As a Student Analyst, I gathered and analyzed technical data on a number of social issues that the Center was targeting for research. I also designed reports displaying this data and interpreted the data to identify key metrics and draw conclusions.