SESSION

NextGen Medical Affairs: Medical Response and Literature Review with GenAI

OVERVIEW

EXPERIENCEIn Person
TYPEBreakout
TRACKGenerative AI
INDUSTRYHealth and Life Sciences
TECHNOLOGIESAI/Machine Learning, Delta Lake, GenAI/LLMs
SKILL LEVELIntermediate
DURATION40 min
Client: Vertex PharmaceuticalsThe client had access to rich, disparate, unstructured data sources. Insight identification was manual and labor-intensive to identify patterns in HCP sentiment.Build an NLP platform that uses unstructured data from sources like Scopus, Pubmed, and Snowflake to provide insights. The platform extracts data using the Medallion architecture, feeds multiple models, and generates reports based on use cases. The platform also performs feature engineering, leveraging JSL & LLM models for key word extraction and text summarization. Visualization of text using NLP, JSL, and LLM models is being developed for improved searchability and relationship identification in a question-answer bot.Value: NLP models in Databricks are utilized to cleanse and maintain data from MSL and medical sources, providing valuable insights for client strategy and medical affairs teams.

SESSION SPEAKERS

Mariana Lipi

/Director of Analytics and Innovation
Vertex Pharmaceuticals

Abhijay Datta

/Head of Delivery - HLS
Tredence