AI meets chemical engineering

Prof Alexei Lapkin, who leads the Sustainable Reaction Engineering group at the university, is part of a new £800k project developing new AI methods to accelerate drug development.

The Alchemy Frontier Fund are funding the project – supported by industry leaders AstraZeneca and a start-up Chemical Data Intelligence – to focus on developing new AI tools. The project will demonstrate the new methods in trying to overcome a major hurdle in the pharmaceutical industry of scaling up. Promising results in controlled lab settings are often difficult to translate into reliable production processes that meet strict quality and regulatory standards, but it is hoped that the introduction of specialised AI agents will support human experts in process development for manufacture of medicines.

The team believe this would help get impactful medicines out of the lab and into clinics and hospitals faster, resolving the difficulty of bringing complex chemical manufacturing into practice.

Alongside Dr Antonio Rago (King’s College London) and Prof Francesca Toni (Imperial College London), Alexei’s team based at the Department of Chemical Engineering and Biotechnology (CEB) in Cambridge will develop next-generation AI tools to support chemists and engineers in designing and optimising pharmaceutical processes.

The approach centres on building systems that not only generate solutions, but make their reasoning transparent and open to scrutiny. By combining computational models with expert chemical knowledge, the project aims to improve how complex chemical processes are designed, tested and translated into practice.

At its core is a new class of AI ‘agents’ designed to explore and compare different process options in a structured way. Working in multi-agent systems, they generate and test possible process designs, but are guided by established chemical knowledge and continuously checked against human expertise.

This ‘human-in-the-loop’ approach ensures the tools remain focused on real-world chemistry, using expert knowledge to shape and validate their outputs rather than replacing it.

Prof Lapkin said: “These systems are not about replacing expert judgement, but about extending it. By integrating computational tools with chemical engineering knowledge, we can explore process options more systematically and improve confidence when moving from laboratory results to industrial scale.”

Over the next two years, the team will build and test these AI tools in pharmaceutical applications, alongside wider areas of chemical manufacturing and materials processing.