Amsterdam's Cradle raises $73m Series B from IVP to sell protein-design AI to more labs

Cradle, the Amsterdam-based maker of AI software for protein engineering, has raised $73m in a Series B round led by the US investor IVP. Index Ventures and Kindred Capital, both already on the cap table, put in more money, and the company says its total funding now tops $100m.

The size of the round matters because Cradle only started charging for its platform at the beginning of 2024. Within that first year of sales it signed more than 21 customers, who between them are working on 31 proteins with its tools. The new round is meant to turn that early traction into a much wider customer base.

What the software does

Proteins sit inside a long list of products: drugs, vaccines, enzymes for detergents, food ingredients, crop protection and industrial chemicals. Improving one, for example making it more stable or more active, has traditionally meant years of trial and error at the bench, with no promise that the target will ever be reached.

Cradle's approach is to pair generative models with the customer's own experimental data. Scientists use the platform to propose new protein sequences, test the most promising ones in the lab, and feed the results back. Each round teaches the models more about what works for that protein, so teams should need fewer cycles of experiments to get where they want.

The company puts numbers on the effect. It says customers have finished projects between 1.2 and 12 times faster and cut costs by as much as 90%. Those figures come from Cradle itself and have not been independently audited.

Its client list already includes some large names. Novo Nordisk, Johnson & Johnson Innovative Medicine, Novonesis and Grifols all use the software, according to the company, and over the past year it also announced partnerships with Novo Nordisk and Ginkgo Bioworks. Customers come from pharma, chemicals, food, agriculture and materials.

Where the $73m goes

Cradle has named three priorities for the new money:

• A larger wet lab. Running its own experiments lets Cradle produce fresh datasets on additional protein types and properties, which it uses to train its foundation models.
• More engineers. The machine learning team will grow so the models can handle more complex proteins and generalise across a wider range of tasks.
• Sales and operations. The company wants to get the platform into many more scientific teams.

The commercial push already has a leader. Sam Partovi joined as chief commercial officer, bringing experience in scaling life science platforms; he was the 40th person on the payroll, which gives a sense of how lean the company still is relative to the capital it now holds.

Co-founder and chief executive Stef van Grieken, who previously worked as a product manager at Google, said the past two years of research and partner work had shown results in areas ranging from vaccines and diagnostics to sustainable chemicals and crop protection. He set a blunt target for the next phase: putting Cradle's software in front of a million scientists. Van Grieken started the company with Daniel Danciu, who serves as chief technology officer.

Why a US growth investor came to Amsterdam

IVP is a North American investment firm that runs a fund dedicated to European technology companies. Its choice to lead this round signals that investors see protein design as a software business that can be sold widely, rather than a services arm attached to a drug pipeline.

Alex Lim, a general partner at IVP, framed the bet in broad terms.

"Biology is one of the domains where generative AI can have the biggest positive impact," Lim said.

He added that any savings at the research stage in fields like drug discovery would translate into large financial returns for customers, and described Cradle as one of the most consequential AI companies in Europe and beyond.

For European founders, the deal is a useful data point. Much of the attention in AI and biology has gone to companies that design molecules in-house and take them into clinical trials themselves. Cradle has chosen the other path: it does not own the products, it sells the tools to the companies that do, across sectors where a protein is the core ingredient. That model spreads risk across many customers, but it also means growth depends on persuading cautious R&D teams to change how they work.

What comes next

The test for Cradle is whether early adopters turn into broad deployments inside large companies, and whether the wet-lab data it generates keeps its models ahead as more groups chase the same problem. The company has kept adding partners since the round, announcing collaborations with Hansa Biopharma and, most recently, with UCB on antibody engineering. Its stated goal of a million scientist users remains a long way off from the couple of dozen customers it reported when the Series B closed.