Vrunda Bhattbhatt

Machine Learning

Successfully handled all content and resource arrangements for Hackathon in Atlanta for Analytics product in Varian with customers in America. Volunteered as a focus group to derive problems and suggest solutions for Glint Survey in Varian. Created a model with dice score of 0.74. Working on OCR projects to read characters on different surfaces like Tires and rubber belts. Work on new POCs for OCR projects. Guide team to improve accuracy for OCR projects using different techniques. Created and deployed Machine Learning pipelines using kubeflow. These pipelines helped in reducing time taken to train semantic segmentation models and tune hyperparameters. Drive functional and technical requirements of AI Platform team. Conducted extensive research on medical image segmentation and improved model for liver & liver tumor segmentation. Conceptualize and developed deep learning model for RT Dose Prediction for AAPM challenge that helped in improving score of dose prediction. Constitute classification model for patient attrition prediction. Used K-means clustering to inference bad data markers in machine logs that is used to predict machine failures. Significant improvement in liver and liver tumor segmentation demonstrated using deep learning techniques. Lead engineering team for hackathon with customers in USA to build customized dashboards from existing data models to generate more revenue from the existing product. Extensively worked on ETL performance improvement for EHR analytics product. Designed datawarehouse schema for Medical Oncology operational data. Groomed team of 7 and helped the remote software deployment team implement Agile methodology. that improved team's performance significantly and team received innovation of the year award for the product developed during my leadership. Gathered and defined customer requirements to develop clear specifications for creating well organized project plans. Develop and web application focused for family communications. Le