Worked under Dr. T K Srikanth and Dr. Ramesh Kestur at International Institute of Information Technology Bangalore. The aim of the project was to research and develop a deep learning model implemented using a novel version of UNet architecture to detect necrosis by semantic segmentation. I was also involved in the creation of a high-quality ground truth dataset was a secondary task through data manipulation and augmentation techniques along with performance analysis. The new model surpassed the previously existing Conditional random field (CRF) model by 13.724% in the AUPRC metric. The work done in this project is in the final stage of review. It is set to be published in Elsevier’s ”Data in Brief” Journal soon, and I am a co-author of this paper.

Developed a Speech-to-Text system for digits by using digit utterances and a handwritten Telugu character recognition system under Dr. Hema Murthy at Indian Institute of Technology Madras. Utilized ML techniques such as Dynamic Time Warping (DTW), Hidden Markov Model (HMM) and Artificial Neural Networks (ANN) to achieve high accuracies in classification.

Worked under Mr. Sunil Agarwal at Wells Fargo and Company for the development of a production quality Trade Order Matching Engine capable of performing matching of orders extremely fast. It was modularized into UI, server, matching engine and database. Worked primarily on the back end of the engine and created a novel data structure (based on Red-Black trees) that enables matching of orders (and other operations such as insertion and amends) to be performed in O(1) time.