Assisted 60 students in understanding and completing computer vision labs and homework via office hours. We did a lab for Hough Transform, Image Segmentation, Image Stitching, QR Codes, Image Classification via Deep Learning.

Trained XGBoost Regression model on 400K records in S3 to estimate the duration of language translation projects. Created a Machine Learning Pipeline to continuously clean data, train a model, evaluate that model, and deploy it.

Coached students through competitive coding questions on HackerRank to prepare them for technical interviews.

I examined the natural phenomenon of Taylor's Law in population growth and migration over time. It turns out that this phenomenon has been observed in a wide variety of life forms from plankton to HIV clusters. I sought to understand and prove the mathematical forces that drive this asymptotic behavior. This involves using differential equations and a population growth and migration matrix to represent several sub-populations in a network. As our understanding of Taylor's Law progresses, biologists/ecologists will be able to take into account migration between sub-populations when creating models of species population growth/decay.

Built computer vision models to detect empty and occupied containers, read container labels, and track inventory over time. This involved digging into the Detectron2 library in Python. Used these models in a Jetson Nano inside the warehouse to perform the above tasks.