A better patient experience begins with listening to people when they reach out.

This session presents recent work done at Kaiser Permanente with John Snow Labs to apply state-of-the-art natural language processing to identify patients’ key concerns, clinical severity, and emotional state from customer digital engagements.

It focuses on what it takes to move such solutions from a proof of concept to a production system: a secure & compliant enterprise data science platform; feature extraction and engineering, model governance including experiment tracking, versioning, and models as code; and scalable “pre-train & tune” NLP workflows.