Episode 4: Acceleration of Process Development & Characterization With Hybrid Modeling & Transfer Learning

Biopharma Developing manufacturing processes for therapeutic antibodies more efficiently depends on making better use of existing process knowledge. Hybrid modeling and transfer learning are emerging as powerful approaches that enable knowledge to be reused across molecules, processes, and development stages, helping to accelerate antibody process development while reducing the need for extensive experimentation. In Episode 4 of the Therapeutic Antibody Series 2026, Moritz von Stosch will explore how these machine learning-driven approaches enhance process understanding, reduce experimental requirements, and shorten development timelines while supporting more robust process characterization. This session will demonstrate how advanced modeling strategies are helping to streamline therapeutic antibody development, from early process optimization through to manufacturing. Key Learning Objectives: How hybrid modeling combines mechanistic and data-driven approaches to improve antibody process development How transfer learning enables process knowledge to be reused across molecules and development stages How advanced modeling strategies reduce experimental effort, strengthen process characterization, and accelerate manufacturing timelines Moritz Von Stosch, PhD, HINA Bioventures SinoBiological