Forum Co-Chairs: Natalie Ciaccio, Vir Technology and Marc Verhagen, Sanofi
The generation of technical data to support the development and manufacture of pharmaceutical products and associated regulatory submissions is fundamental to the pharmaceutical industry. The recent boom in artificial intelligence (AI) systems and large language models (LLM) has resulted in increased interest in the capability of these systems to support industry activities. By embracing technological advancement, the biopharmaceutical industry can make significant strides in reducing submission timelines and accelerating the delivery of therapeutics. This session will include discussion on current practices with examples of how structured data, digitalization, AI, and generative AI (GenAI) are being leveraged to support pharmaceutical development, manufacturing, and standardized regulatory authoring/submission with dynamic review. The focus will be on opportunities to improve efficiency or accelerate development timelines. Discussion will include considerations for emerging technologies and forward-looking approaches for analytical data modeling.
Different aspects of the use of AI or LLMs during CMC development will be discussed including:
• Use of AI in drug discovery/developability – modeling of protein structure and properties based on sequence
• Bioprocess/ cell culture optimization using predictive analysis
• Stability modeling using AI and/or machine learning
• Leveraging commercial manufacturing data/modeling to optimize process conditions
• Automated generation of regulatory documents using LLM
Session Speakers:
The Challenge and Criticality of Data Contextualization to Support Effective AI/ML/Modeling in Bioprocess Development – an Industrial Case Study
Christian Airiau, Sanofi
Development of Realistic and Safe AI/ML Applications for Biotherapeutic Characterization and Process Development
Jeremy Shaver, Pfizer, Inc.
A “One-Click” Submission with AI and Digital Authoring
Kabir Ahluwalia, Amgen Inc.
Practical and Regulatory Considerations for Machine Learning Models Applied to Process Development and Control
Ben Stevens, GlaxoSmithKline