Getting Pharma AI Agents to the Right Answer by Fixing the Data and Semantic Layers - with Richard Knoche of Flatiron Health

Pharma is moving toward AI agents, but agents only know the context they are given, so gaps in data and meaning turn into wrong answers at scale. In this episode, Richard Knoche, Head of Data Science and AI at Flatiron Health, explains why a strong data foundation and a well-built semantic layer determine whether AI agents can be trusted with complex real-world data. He walks through how historical human analyses can be used to build and benchmark that semantic layer, why having agents resolve ambiguity before answering improves accuracy and reduces handoffs between teams, and why leaders need centralized data governance and agents that plug into any system rather than siloed, bespoke tools.

​

This episode is sponsored by Flatiron Health

​

Emerj works with a select group of AI vendors to reach Fortune 500 decision makers through research, media, and direct access. If you want to be considered, download our media kit at emerj.com/AD1