738: Engineering Biomaterials with Generative AI — with Dr. Pierre Salvy

Introduction

Dr. Pierre Salvy, head of engineering at Cambrium GmbH, is pioneering the use of generative AI in material design, specifically in creating biological proteins for diverse applications. At Cambrium, he has expanded the team to bridge the gap between wet lab biology, data science, and scientific computing, using large language models to innovate in biomaterials. Pierre's work has led to the rapid development of NovaColl, a vegan collagen, within just two years, showcasing the potential of AI-driven research and development 1. He explains, "We use generative AI to create biomaterials for tomorrow," emphasizing the transformative potential of this technology 2.

   

Technology

The technological processes at Cambrium involve a sophisticated blend of genetic engineering and machine learning, described as a "double deep tech sandwich" by Pierre. This approach utilizes generative AI to optimize protein sequences, akin to how natural language processing advances have improved language models 3. Pierre elaborates on the analogy between protein and natural language, stating, "Proteins are basically sequences of letters," highlighting how AI can generate functional proteins from vast possibilities 4. This innovative process allows Cambrium to design proteins with specific functions, revolutionizing material science.

   

Impact

Generative AI's application in creating vegan collagen, such as NovaColl, demonstrates its potential to transform industries like cosmetics and plastics. Pierre discusses the challenge of producing animal-free collagen, which is traditionally derived from animals, and the innovative solution Cambrium offers 5. Beyond cosmetics, Cambrium aims to replace plastics with sustainable materials, leveraging AI to explore new protein-based materials 4. "We want to be a company that replaces plastics in the future," Pierre asserts, underscoring the broader environmental impact of their work.