About

Amit Kumar is a diligent and skilled data scientist with over 6 years of experience in the field. He possesses strong communication skills and a solid background in data science, data visualization, and machine learning. Amit has hands-on experience in deep learning and data analytics projects, making him a valuable asset to any team. Amit's diverse background provides him with a unique profile that combines his expertise in evaluating state-of-the-art machine learning algorithms and advances in DL software frameworks. He has worked on NLP use-cases such as sentiment analysis, building question-answering systems, and speech-to-text models. Amit has also built rigorous deep learning technologies applicable to the computer vision domain, such as object detection in videos and scene classifications. Amit has the ability to work on detailed technical tasks with the vision for the big picture. He can create effective visualization charts from studying data and work in a team with good organizational skills and a positive attitude. Amit is always open to networking and interested in connecting with like-minded individuals. Currently, Amit works as a data scientist at LogicPlum, where he reviews computer vision problems predominantly as part of Kaggle competitions and works on reproducible code, thereby applying to other public datasets. He has worked on LogicPlum Vision, part of the company's AI Innovation platform designed to automate the image classification process, including producing ML models in a short turnaround, thereby decreasing the inference time. Amit has also worked on ML fairness, ensuring that biases in the data and model inaccuracies do not lead to models that treat individuals unfavorably based on characteristics such as race, gender, disabilities, and sexual or political orientation. Amit has previously worked as a data scientist at Deski, where he conducted a literature review of medical journals on the use of deep learning in view classification of echocardiography images and applied those models in the project. He also worked on Xplainable AI to understand model interpretability, developed custom loss functions to train models for better performance, and applied various pre-processing and data augmentation techniques to reduce the inference time for view classification projects and improve results. Amit has a Master's degree in Data Science and AI from the Data ScienceTech Institute and a Bachelor's degree in Electronics & Telecommunication Engineering from CV Raman College of Engineering. He possesses expertise in several technical skills, including statistics, data analytics, and visualization, machine learning, deep learning, computer vision, NLP, Elastic Search, Grid Computing, Unix, and SQL.