Cradle is a Dutch-Swiss biotechnology company that builds artificial intelligence software to help scientists design and optimize proteins with greater efficiency. Founded in 2021 and headquartered in Amsterdam with an office in Zurich, the company develops machine learning models that learn directly from experimental data and generate protein sequences predicted to exhibit specific properties. This automated design process reduces reliance on trial and error while allowing research groups to test broader sets of hypotheses within shorter timeframes.
Traditional protein engineering often depends on sequential experimentation, where one property is optimized before another is considered. That method can consume months of laboratory work and substantial resources. Cradle integrates computational prediction with laboratory feedback so that multiple protein characteristics, such as stability, activity and manufacturability, can be evaluated together. As researchers upload assay results into the system, the models refine predictions and propose improved variants that reflect real world findings.
By grounding predictions in experimental evidence, the platform connects digital modeling with biological reality. Scientists remain in control of experimental direction, yet the software proposes options that might not surface through manual exploration alone. This iterative loop between computation and laboratory testing reshapes how proteins are discovered and refined.
Accelerating Discovery Through Iteration
Cradle relies on generative artificial intelligence trained on extensive protein sequence datasets, which are continuously refined with user-specific experimental results to produce more precise predictions. Instead of optimizing a single trait, the system evaluates multiple objectives simultaneously and generates candidate sequences that balance competing requirements. Researchers can rank these candidates according to predicted performance and select the most promising variants for synthesis and laboratory validation.
This method allows laboratories to allocate resources more strategically. Rather than screening thousands of random mutations, scientists focus on variants with higher predicted probability of meeting project criteria. As each experimental round concludes, new data feed back into the model, sharpening subsequent predictions and expanding the searchable sequence space. Over time, the software becomes more tailored to the biological system under investigation.
The platform also presents visual analytics that summarize how predicted variants compare across performance metrics. These visualizations assist researchers in interpreting results and planning next steps, while maintaining traceability between computational suggestions and laboratory outcomes. By structuring information in an accessible format, the system supports rigorous documentation alongside innovation.
Enterprise Adoption and Growth
Since its founding, Cradle has attracted significant venture capital backing that reflects growing interest in AI driven biotechnology. Funding rounds have enabled expansion of research operations, software engineering and customer support across Europe and North America. Adoption among pharmaceutical and biotech organizations demonstrates that computational protein design is moving from experimental concept to operational tool.
Large drug developers face long development cycles and high costs when optimizing biologic therapies. Tools that generate improved protein variants with fewer laboratory iterations can reduce timelines from early discovery to lead candidate selection. Cradle’s enterprise grade platform integrates into existing research infrastructures through secure data environments and application programming interfaces, ensuring compatibility with established workflows.
Security remains critical in drug discovery and industrial research. The platform incorporates encryption and controlled access mechanisms to protect proprietary information, allowing organizations to retain ownership of experimental data while benefiting from machine learning insights. This architecture supports compliance with regulatory requirements and enterprise standards.
Applications Across Industries
Protein engineering extends beyond pharmaceuticals. Agricultural biotechnology firms design enzymes and proteins that influence crop resilience and nutrient utilization. Industrial biotech companies develop catalysts for chemical synthesis and sustainable manufacturing. Food technology innovators explore alternative proteins with targeted texture, flavor and nutritional profiles. Across these sectors, optimization of molecular properties can determine commercial viability.
AI assisted design offers researchers the ability to examine larger portions of sequence space than manual experimentation would permit. By predicting how specific amino acid changes may alter performance, scientists can refine molecular candidates with greater precision. In agriculture, this may involve designing proteins that function effectively under specific environmental conditions. In materials science, it may involve tailoring enzymes that operate efficiently within industrial processes.
The ability to iterate rapidly supports experimentation that once required extended laboratory cycles. Computational suggestions guide laboratory validation, and validated data refine computational models in return. This reciprocal system reduces uncertainty and expands scientific exploration.
Building a Programmable Biology
Artificial intelligence now plays a growing role in life sciences, yet successful deployment requires integration with empirical research. Cradle maintains internal laboratory capabilities to validate model outputs against experimental results, ensuring that predictions align with measurable biological behavior. This linkage between software and wet lab testing strengthens reliability and supports trust among researchers.
Challenges remain in generalizing predictive models across diverse protein families and experimental conditions. High quality data are essential for robust machine learning performance, and continuous validation is required to maintain predictive accuracy. Nonetheless, the convergence of artificial intelligence and molecular biology marks a pivotal shift in how proteins are engineered.
As organizations seek more efficient discovery pipelines, tools that shorten experimental cycles while preserving scientific rigor gain relevance. By enabling data driven protein design, Cradle contributes to a broader movement toward programmable biology, where digital systems guide molecular innovation and laboratory insights refine computational understanding. Through this iterative synergy, biotechnology research enters a new phase defined by integration rather than isolation, and by acceleration grounded in empirical evidence rather than guesswork.
Stef van Grieken, Co-Founder & CEO, Cradle
