Foundation Models and the Future of Bioengineering
Engineering biology is undergoing a rapid transformation, driven by the convergence of AI, advanced modelling, and the expansion of biological datasets. At the centre of this shift is Ian Taylor, Head of Bioengineering at Cambridge Consultants, part of Capgemini Invent .
With more than twenty years’ experience across biomanufacturing, enzyme engineering, therapeutic peptide production, DNA data storage, and next-generation manufacturing technologies for cell therapies, Ian has spent his career at the forefront of biotech innovation. His work now focuses on building AI-enabled platforms that combine foundation models, mechanistic modelling, and digital twins to make biology predictable, scalable, and commercially impactful.
Ahead of his talk at SynbiTECH 2025, we spoke with Ian about the role of foundation models in engineering biology, the opportunities created by AI-driven predictability, and how collaboration and translational expertise can accelerate progress from concept to real-world application
What will you be exploring in your talk “Foundation Models – Unlocking Predictability and Scale”, and why is this topic important for the engineering biology community?
“Biology is very complex. The emergence of tools like AlphaFold have shown the utility of Transformer architectures used in Large Language Models (LLMs) to help unravel some of the complexity and nuance in nature’s rules. The latest protein LLM-based tools enable us to generatively design proteins of desired function. This level of capability was unimaginable a few years ago”.
How are foundation models changing the way we approach prediction and design in engineering biology?
“Foundation models represent a core understanding of the underlying science or phenomena being studied; they provide a framework in which to contextualise application-specific results. This enables the in silico simulation of new design hypotheses which can then be tested experimentally. Critically, this reduces effort in the lab and often yields good results with surprisingly low effort compared to traditional methods. Foundation models are an accelerator in engineering biology”.
What do you see as the biggest opportunities for applying AI-driven predictability in biotech and engineering biology?
“Ultimately, it’s simulation of biological response and design at genome and population scale. This might be the design of an industrial microbe used in the production of sustainable chemicals or designing novel interventions in gut microbiome to affect wellbeing and long-term health”.
In your view, what distinguishes a foundation model from other AI approaches currently used in life sciences?
“Foundation models are useful for capturing broad patterns in biology, but they’re only one part of the toolkit. Mechanistic and physical models are also important and complementary, such as flux balance analysis of cell metabolism or computational fluid dynamics describing bioreactor mixing behaviour. The real value comes from combining these approaches: using foundation models alongside mechanistic and physical simulations to better design, predict, and control complex biological systems like microbial bioreactors”.
What kinds of challenges do companies face when trying to integrate AI or data-driven tools into their R&D workflows?
“A consistent challenge is the lack of structured data. Legacy systems often use incompatible ontologies, so standardisation becomes essential. For example, companies need instruments and software that can communicate seamlessly, ensuring every device connects directly to an ELN rather than relying on a scientist with a USB stick. There is no single solution; describing data schemas in a focused protein engineering workflow is far easier than in a broader, more heterogeneous R&D environment with a wider set of data types and a more exploratory approach, like a skin science lab within a consumer goods company”.
Capgemini works at the intersection of science, technology, and commercial innovation, how does that position you to lead in this space?
“We have privileged insights into our clients’ challenges which drives their need for continuous innovation to maintain best in class performance. Our transdisciplinary teams of strategists, scientists and technologists across Capgemini and our deep tech powerhouse, Cambridge Consultants, enables us to create breakthrough technologies, often starting from first principles, with a clear focus on delivering commercial impact. We partner with visionary organisations, applying advanced capabilities and industry expertise to help them achieve their business goals. Our approach is always bespoke, tailored to each client and their specific business needs”.
Collaboration is a major theme at SynbiTECH, what role do partnerships play in accelerating progress in this field?
“The technology stack in AI-enabled engineering biology is incredibly deep, and very few organisations have every capability in-house. Our focus is on solving real client problems and we build a partner ecosystem that brings the right expertise to each engagement. Collaboration becomes a force multiplier, unlocking progress faster and delivering outcomes that matter”.
What excites you most about the future of predictability and design in engineering biology?
“What excites me most is how close we’re getting to true de novo design of microorganisms, building new biological systems and reliably predicting how they’ll behave at scale. A convergence of technologies is emerging, and I expect Agentic AI will play a central role in orchestrating these capabilities. Together, they have the potential to reshape how we design, test, and understand biology, opening the door to breakthroughs that were previously out of reach”.
Finally, why is Capgemini proud to support SynbiTECH2025, and what do you hope attendees take away from your session?
“SynbiTECH 2025 showcases the very best of the UK’s engineering biology ecosystem, bringing together its dynamic SME and academic communities to drive the bioeconomy forward. In our session, I hope to spotlight the vital role that professional services and technology providers play in turning promising ideas from this community into world-changing solutions. Ultimately, it’s about helping our clients harness these innovations to solve real problems and accelerate impact”.
As SynbiTECH2025 brings together innovators from across the UK’s bioeconomy, Ian’s insights underscore the importance of foundation models and agentic AI in unlocking the next generation of sustainable biomanufacturing, advanced therapeutics, and progressing de novo biological design. Ian’s session promises a compelling view into how AI-guided science can reduce uncertainty, accelerate development, and empower organisations to scale with confidence.
