Thomas R. Gardos

Since Fall of 2023, I am Associate Professor of the Practice with Boston University’s Faculty of Computing and Data Science and the inaugural director of the MS in Data Science program and inaugural director of the CDS Clinics program.

My interests are large multimodal and language models, computer vision, neural networks and data science more broadly. My approach is to blend theory and practice both in my teaching and also research projects. Theory enables you to solve problems from first principles if necessary and practice often uncovers new problems that require better analytical approaches. Where possible, I like to deploy solutions as open source projects.

I teach Deep Learning for Data Science, ds542, which I introduced in the Spring of 2024. I also teach the graduate level course Tools for Data Science, ds701. I’ve also taught ds210, the undergraduate Rust programming course and have taught (led, coached) the senior-level BU Spark! Machine Learning Practicum over numerous semesters.

Prior to joining BU, I spent 30 years at Intel Corp., most recently leading a small R&D team developing low-power computer vision algorithms and contributing to the definition of an ultra-low power computer vision ASIC.