About

Uday Khanna is a highly skilled data scientist with almost 4 years of experience in statistical analysis, machine learning, data visualization, and creative thinking. He is passionate about coding and solving complex business problems using his deep statistical background. Currently, he works as a Data Scientist 2 - NLP and CV at Episource LLC, where he is responsible for textual entailment using deep learning, training models on AWS Sagemaker via Docker, and analyzing predicted data for false positives/false negatives. He is also involved in identifying tables from medical records using CV and using that information for other tasks. Prior to his current role, Uday worked as an NLP Engineer at Cunesoft GmbH, where he built custom named entity recognition models to identify drug products using deep learning. He also built a text classifier using machine learning and regular expressions to identify the type of regulatory document and then extract the required information. He automated the process of extraction of specific dates and manufacturers' addresses and responsibilities along with the drug substance. Additionally, Uday built an image classifier to identify health authorities and automated the process of extraction of information using OCR and pdfminer. Uday also worked as a Data Scientist at Dhwani Rural Information Systems, where he performed SQL queries on MongoDB using Spark SQL and designed a pipeline for a live streaming dashboard using seaborn, socket io, and pymongo. He visualized data using matplotlib, Tableau, PowerBI, and seaborn. Uday holds a Bachelor of Science in Statistics from Delhi University and a Post Graduate Diploma in Data Science from IMS Pro School. He is highly skilled in NLP, research scientist, data scientist, computer vision, SQL, deep learning, ML, web, Sagemaker, Python, AWS, Flask, BERT, Docker, MongoDB, Azure, Seaborn, information retrieval, and Spark. With 6.86 years of relevant experience, Uday is seeking data-related problems to solve and is committed to delivering high-quality results.