About

Pragya Sahu is a Chief Engineer at Samsung R&D Institute India with a specialization in Microelectronics and VLSI Design. She holds a B.Tech. in Electronics and Electrical Communication Engineering from the Indian Institute of Technology, Kharagpur, and a dual degree M.Tech. in Microelectronics and VLSI Design. Currently, Pragya is working as a Chief Engineer (Research) in the Advanced Technology Labs at Samsung R&D Institute. Her work focuses on implementing deep learning methodologies for sequential learning, with a special focus on handwriting recognition. Pragya has also worked in the field of Natural Language Processing (NLP) as a part of the domain classifier system for Samsung Bixby. She is interested in developing smart and intelligent solutions for challenging problems. Pragya has over 5.5 years of relevant experience in the field of Research and Development. She has previously worked as a Lead Engineer at Samsung, where she developed an online handwriting recognition system for Hindi, Marathi, and Hinglish. She built a B-LSTM with CTC cost function based online handwriting recognition module and developed language-specific processing techniques to improve character recognition accuracy. Pragya also implemented a novel Hindi-specific space analysis technique that improved word recognition accuracy. The solution was released in Samsung Note-8 devices. Pragya has also developed and implemented a comprehensive system for recognizing and completing partially written text, using the natural writing way to obtain additional information of required word length and built a hybrid ranking system using n-gram context, word frequency, and required word length. This project was patented and the paper submitted in Mobile-Human Computer Interaction - 2019. The paper won the Samsung Best Paper Award - Silver at a global level. Pragya has also worked on projects handling phonetic word variations in Hinglish text recognition and on-device personalization of trained models focused on handwritten text recognition. Additionally, she has worked on domain (App) classification for natural language commands for Samsung Bixby. Before joining Samsung, Pragya worked as a Global Technology Intern at Barclays Technology Centre India. She worked on automating the weekly report generation using the concepts of machine learning, decision tree algorithms, and neural networks to design an algorithm for reading the database and deciding the sub-clauses. Pragya's tech stack includes Research Scientist, Mobile, Deep Learning, NLP, CNN, n-grams, Senior Software Engineer, ML, and Neural Networks