Hello, I am Utkarsh, master's student in Robotics and Artificial Intelligence at Arizona State University, with a strong foundation in Computer Science.

My research experience is extensive and varied. I first worked at the Indian Space Research Organisation (ISRO), where I developed an algorithm using Convolutional Neural Networks (CNNs) to segment buildings in aerial images. This project involved achieving a 95% Dice Score and implementing shadow detection algorithms, which improved image segmentation accuracy by 5% using MATLAB and Python.

Subsequently, I worked at the Indian Institute of Technology (IIT - BHU), where I contributed to developing a Neural Machine Translator (NMT) for Indian languages based on Transformer Neural Networks. I integrated a positional encoder, which improved translation accuracy by 15%, and researched Monotonic Chunk-wise Attention, enhancing model performance by 3% on supercomputers.

At ASU, (in my ongoing project) I implemented TRPO and SAC algorithms to enhance Safe Explicable Planning for robots, ensuring safety and aligning AI decision-making (application of reinforcement learning) with human expectations.

Professionally, I served as a Software Engineer in Data Science and Machine Learning at BusinessNext, collaborating with over 10 clients to enhance CRM software with OCR and Forecasting Leads Models, increasing customer analysis accuracy by 20%. I utilized MS SQL Server to improve data processing efficiency by 30% and applied machine learning algorithms for predictive analysis.

I am proficient in several programming languages and tools, including Python, TensorFlow and more. My expertise spans AI, ML/DL, Reinforcement Learning, NLP, Computer Vision and GAN. Additionally, I am passionate about sharing my knowledge for everyone's benefit. I believe that knowledge should be accessible and used for the greater good. This drives me to write blogs on Medium (you can check my blogs by going to the Home section).

Resume

Email

LinkedIn

GitHub

X