• Implemented machine learning models from scratch on public as well as internal datasets of chemical compounds to screen for drug discovery using Scikit-learn and Pytorch• Adapted open-source machine learning models for customized datasets and scaled models using Ray on Anyscale• Deployed a Streamlit app for custom data uploading, processing and ingestion for inference based on regression and clustering models• Curated custom datasets by programmatic API queries for training and evaluation • Created linked data models and schema for preparing knowledge graph and graph database of clinical trials, research publications, diseases and drugs

Built and deployed ML pipeline: AWS Sagemaker, NVIDIA edge devicesManipulated DICOM datasets to perform image segmentation: python, pydicom, scikit-imagePerformed meta-analysis from clinical data and generated reccommendation for design decisionsAssembled a cross-functional team and perfomed security risk assessment for HIPAA compliance: HHS SRA toolkitDerived user needs and design input requirements, collaborating with marketing and regulatory teams: Jama Connect

• Discovered significant relationships between electrical properties underlying heart rhythms and structural changes in heart diseases• Developed a novel system and protocol to apply current and measure 3D electric field distribution for application in heart tissue

• Product Development in Body Electronics of SUVs and MUVs• Development of in-vehicle and automated test cases for body electronic functions