Webinar: Enhancing Medical and Product Insights with Automated NLP - Tom Horan
Tom Horan described how the Scientific Analytics group in Novo Nordisk extracts actionable insights from internal and external data sources by leveraging tools like natural language processing, machine learning and predictive analytics. The team creates automation workflows and exposes insights through dashboards using visualization applications like Tableau and Qlik.
Novo Nordisk uses its Amazon Web Services data lake to host multiple data sources and it has reduced ongoing costs and manual effort by automating and reusing workflows and standardizing the appearance and communication of insights via dashboards. As examples, the automation pipelines developed for a Covid-19 project were then applied to an obesity publications impact tool; and machine learning techniques were used to develop key opinion leader mapping and patient recruitment modeling tools.
Tom drilled into the underlying technology and automation workflows involved in the Medical Insights Dashboard, which is used to monitor and explore questions from healthcare professionals coming into the Medical Information Department and medical affairs teams. The questions are text mined with Linguamatics NLP, and KNIME is used to build the pipeline, schedule workflows to acquire data from the OASIS data lake, parse the data, and connect the jobs and scripts. The results are made available in the Dashboard via Tableau.