Staff Data Scientist
Current• Trained a BERT style model from scratch, utilizing internal natural language data on an 8 GPU machine on Amazon Web Services (AWS). • Developed a vectorized search engine to migrate existing terms to ICD-10 5th edition codes. • Designed and productionized a Named Entity Recognition model to extract medical problems and procedures from medical free text. • Advised two data science master students in a capstone project focused on reorganizing Intelligent Medical Objects data, conveying detailed information and providing comprehensive mentorship. • Leveraged large language models (LLMs) to bolster internal model performance, effectively improving name recognition and concept mapping for critical applications. • Collaborated with several physicians serving as a machine learning subject matter expert (SME) and delivering support and guidance as needed. • Delivered technical presentations to Chief Medical Officers, physicians, and C-Suite executives, translating complex concepts into easily digestible insights.