Lead Data Scientist
Current- Design and develop deep learning models for a leading home care and hospice electronic health records system, using Python, Pandas, PySpark, Azure, and Databricks.- Fine-tune large language models on clinical outcomes, using PyTorch and Hugging Face. - Develop feature sets to incorporate structured and unstructured data elements. - Extract features from sparse diagnosis and medication codes, using Tensorflow and Keras.- Develop deidentification pipeline for training HIPAA-compliant models. - Won 2023 company hackathon with a prototype of a pipeline for synthesizing overall patient narratives from visit notes, using Langchain and OpenAI ChatGPT.- Preliminary work on optical character recognition / information extraction pipeline using Tesseract and OpenAI ChatGPT.