Skip to main content
RAPS Euro Convergence 2026 Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Presentation Icons
Learning Level: Advanced
Learning Level: Advanced
Learning Level: Expert
Learning Level: Expert
Learning Level: Foundational
Learning Level: Foundational
Learning Level: Intermediate
Learning Level: Intermediate
RAPS Focus
RAPS Focus
Networking Opportunity
Networking Opportunity
Ticketed Event
Ticketed Event
Student Focus
Student Focus
Evaluation ViewAttendees 5

Pharmaceuticals

Documents Are Data You Never Knew You Had: AI Tools Can Unlock Their Knowledge

Wednesday, May 6, 2026
08:30 - 09:30 WEST
Location: Auditorium 4

Description:

Regulatory professionals handle documents every day — guidance texts, assessment reports, scientific advice letters, dossiers. Yet most of them do not think of these documents as data. This session tells the story of what happens when they do.

Three organizations. Three similar contexts. Three innovative approaches to the same fundamental challenge. How to unlock greater value from the accumulated human knowledge captured in regulatory and scientific documents?

- At Bayer, the journey starts with people. In parallel of developing and deploying AI-powered document search and Q&A tools, the team stressed the importance of communication and training. Their experience in change management, use case selection, and upskilling offers every regulatory professional in the room a practical roadmap for getting started.

- At EMA, the challenge is scale and heterogeneity of information. Scientific Explorer uses validated AI prompts to extract specific pieces of information from documents and indexes them as searchable, structured fields. The added value scales with the corpus: a regulator can now query across thousands of scientific advice letters and assessment reports simultaneously, surfacing regulatory precedents with precision that no manual search could match. The knowledge that was locked inside documents becomes navigable, comparable, and actionable.

- At Merck, the ambition is depth. Their project takes the logic one step further: unstructured document content is not just extracted — it is transformed. Captured with aligned vocabularies, enriched with metadata, relationships, and provenance, and stored as FAIR linked data in a semantic knowledge base, it becomes the foundation for agentic AI workflows. The added value is qualitative: the system does not just answer questions about documents — it understands the knowledge those documents contain, enabling AI agents to reason, generate, and act with human oversight.

Three AI technology explorations at the service of regulatory professionals. One shared insight: the value of AI grows in direct proportion to how well we treat our documents as data assets and how well we accompany regulatory professionals in the digital transformation journey.

Learning Objectives:

  • Realize that regulatory documents are a latent data asset that can be leveraged by AI to work more, work faster, and work more precisely for regulatory professionals, which is a trigger for the digital transformation journey.
  • Understand how to unlock regulatory and scientific knowledge and insights from a large corpus of documents with AI tools, based on examples from Industry and EMA that demonstrate different usages of technology, starting with simple and easy implementable solutions for regulatory affairs professionals.
  • Explain how unstructured document content can be progressively structured, enriched with metadata and semantics, and stored as FAIR knowledge base of linked data to power more advanced downstream usage like end-to-end content generation.
  • Identify the human, organizational, and change management prerequisites — including human-in-the-loop design and AI literacy — that determine whether AI tools deliver lasting value.