Worked for a cumulative total of six semesters, including TA'ing both courses in Fall of 2020.

Led recitations for both masters and undergraduate students. Graded assignments, and hosted office hours. .

Developed a prototype lab (Parallel Lab) for 15-213 to evaluate students based on their implementation of a work queue to handle a variety of workload distributions

Developed an office hours queuing system in use for multiple semesters by 15-213. Oversaw a total of 2,000 students and 10,000 questions asked and answered while actively maintained.

Worked on implementation of Fixed Function Acceleration for a low power embedded architecture: MANIC. Worked for a semester long in coordination with an undergraduate research seminar course (15-300).

Implemented units performing fixed functions (sorting, Fourier Transform, 2D strided convolution) using SystemVerilog. Implemented the Eyeriss paper's 2D strided convolution to evaluate performance.

Soterea is a software company that develops an intelligent driving system for commercial vehicles. Worked on site in Tianjin with other interns to develop new machine learning based models to improve monitoring and safety of commercial vehicles.

Developed software with respect to lane segmentation and tracking in a front camera view of a commercial vehicle. Developed stable tracking of lanes using Pytorch and Numpy. Used Kalman Filters to improve performance of noisy samples.

Presented a separate deep neural network composed of various separate networks to fit constraints of performing inference on a computationally limited, embedded platform. Exported the model in ONNX, and used in C++ modules for inference.

Performed data collection, coordinated annotation with dedicated personnel, augmentation, transfer learning, and efficacy of embedding.