About

Rohan Dhupar is a seasoned data scientist with over 4 years of experience in the NLP domain, working with legal, retail, and finance documents. He has a deep understanding of the product development life cycle of data science projects, including developing and deploying machine learning and deep learning models on AWS. He is skilled in data preparation and engineering for training data for machine learning and deep learning models. Rohan has published research on hyperspectral image classification using SWT & CNN in IEEE and hyperspectral image feature classification using GWO-SWT-CNN in Springboard. His skills include Python, DSA, PostgreSQL, machine learning, deep learning, image processing and classification, and new modern age transformers custom model development (BERT/T5) for various use cases from large chunk extraction to large entity classification. He is also experienced in AWS model as micro-service API, Docker, Apache Spark, cloud data warehousing using AWS redshift, and data engineering using Apache Airflow. Rohan has been a guest speaker at NIT Jalandhar 2019, GITAM UNI Vizag (2019), and Kaggle Days NCR (2020). He is fluent in English and German (A1). Currently, Rohan is working as a Senior Software Engineer - Analytics at Airtel, where he is involved in scalable computer vision application development for E-KYC automation/UPC transfer, chat-bot as Hinglish language sales distributor support, live face detection model on 2D still images for detecting fraud during KYC done by retailers, and quantisation of heavy PyTorch/tensor-flow models to TensorFlow-Lite for decreasing response time 4x to 8x times and reducing space by 2x to 4x. Previously, he worked as a Research Scientist (NLP in Legal/Business Domain) at Exela Technologies, where he worked on legal contract understanding and large size chunk & entity extraction using both modern and classic tools of NLP. At OYO, he worked as a Data Scientist (Pricing Anomaly Analytics), analysing pricing and revenue for OYO UK, anomaly detection models for web scrapers, buildings data quality analysis scripts, wrongly mapped properties from different sources detection, web scraper automation bots for various hotel booking websites to build data set for machine learning, cluster analysis of OYO properties vs competitors weekly automated tool in CLI, and UK hotel properties revenue prediction model development. Rohan holds a Bachelor's degree in Computer Science from Raj