In our Next Gen Biomed thought leadership interview series, we explore the ideas shaping the future of life sciences. In this conversation, Stef van Grieken, co-founder and CEO of Cradle, shares how AI is moving biology from trial-and-error to true engineering.

Stef van Grieken, co-founder and CEO of Cradle, is helping lead a new wave of biotechnology—one where artificial intelligence is transforming how proteins, the building blocks of life, are designed.

At the core of Cradle’s platform is a shift away from traditional trial-and-error experimentation toward a more precise, engineering-led approach. “What we’re seeing is biology move from discovery to design,” van Grieken explains.

Reimagining Biological Discovery

When Cradle was founded, the ambition was clear: address the inefficiency and low success rates that have long defined biological discovery.

“Historically, developing new biologics or enzymes could take years, with a lot of uncertainty,” he says. “We wanted to change that by making the process more predictable and scalable.”

Cradle’s AI platform supports both the discovery and optimisation of proteins, enabling researchers to move from early-stage ideas to viable candidates significantly faster. Today, more than 30 organisations—including Novo Nordisk and Johnson & Johnson—are using the platform.

From Trial-and-Error to Intelligent Design

Traditional lab workflows rely heavily on intuition and iterative screening of vast molecular libraries. AI is fundamentally changing that paradigm.

“Our models generate protein sequences based on specific goals—like stability or binding strength—so scientists can focus on the most promising candidates from the start,” van Grieken explains.

This creates a collaborative loop between human expertise and machine intelligence, accelerating development cycles by 2 to up to 12 times.

Where AI Is Making the Biggest Impact

The most immediate applications are in therapeutics, particularly antibody development. But the impact is quickly expanding.

“We’re seeing strong progress in complex formats like bispecific antibodies, as well as in de novo design—creating entirely new proteins from scratch,” he notes.

Beyond pharma, industrial biotech is benefiting from faster enzyme engineering, while peptides represent a promising and still underexplored frontier.

The Data Dilemma

Applying AI to biology is not without its challenges. Unlike digital environments, biological data is often limited, inconsistent, and difficult to standardise.

“Every lab runs experiments slightly differently, which makes it harder to train robust models,” van Grieken says.

Cradle addresses this through a “lab-in-the-loop” approach, where experimental results continuously refine the AI. Even so, bridging the gap between computational predictions and real-world biology remains a key hurdle.

Embedding AI into R&D

For AI to deliver real value, it needs to integrate seamlessly into existing workflows.

“We’re not trying to replace how scientists work—we’re augmenting it,” he explains. “Our platform fits into the tools they already use, from LIMS systems to data platforms.”

This approach is reflected in partnerships with major organisations, including a multi-year collaboration with Bayer, signalling how AI is becoming embedded in large-scale R&D.

A Shift Toward Engineering Biology

Looking ahead, van Grieken sees a fundamental transformation underway.

“As AI reduces cost and compresses timelines, you’ll see much smaller teams building complex biologics in months instead of years,” he says.

This shift could democratise access to molecular engineering, unlocking faster innovation across medicine, food, materials, and sustainability.

In this emerging landscape, AI is not replacing scientists—it is amplifying their capabilities. And as platforms like Cradle continue to evolve, the ability to “program” biology may soon become as routine as writing code.