About

Kush Teppalwar is a highly skilled Data Scientist with a strong aptitude for programming, good reasoning skills, and problem-solving capabilities. He is currently working at AppZen as a Data Scientist 1, where he has been responsible for building various models, including a Text-based Bi-LSTM model for extracting actual items bought for a receipt and their categories with 92% accuracy, an Image and Text-based Binary classifier using CNN and LightGBM respectively to identify Hotel Receipts as Itemized or Non-Itemized to get a high precision of 95%, and a faster-RCNN Inception V2 model to detect bounding boxes around handwritten amounts and dates from Receipts. He has also built a CNN-RNN-Connectionist Temporal Classification based Handwritten Recognition model to recognize handwritten text and a Logistic Regression model to detect the Final amount from Japanese and Chinese Receipts. Kush has a proven track record of delivering high-quality work and is always open to new opportunities. He is a team player who believes in execution and is passionate about using his skills to solve complex problems. Kush has also worked as an Associate Software Developer AI/ML at icertis. He completed his Bachelor of Technology in Computer Science from Pune Institute of Computer Technology in 2020 and his High School Graduation from New Era Senior Secondary School - India in 2016. Kush has a diverse tech stack, including Data Science, Research Science, ML, CNN, Deep Learning, Image Processing, NLP, AI, RNN, Java, Neural Networks, Software Engineering, Hyperparameter, Frontend, Backend, and Python. He has 3.19 years of relevant experience and is always looking to expand his skill set. Kush is an excellent addition to any team and would be an asset to any organization looking for a talented and experienced Data Scientist.