Belgian pharmaceutical company UCB has deployed Cradle's artificial intelligence-powered protein design platform across selected antibody discovery programs, marking a significant enterprise adoption for the AI biotech as competition intensifies in computational drug design.

The partnership, announced November 4, 2026, embeds Cradle's technology into UCB projects spanning hit generation through preclinical candidate preparation. The collaboration centers on a "lab-in-the-loop" workflow where experimental results from each testing round automatically retrain machine learning models, creating a continuous improvement cycle that refines predictions as scientists progress through discovery stages.

Multi-Property Optimization at Scale

Cradle's platform tackles one of antibody development's most persistent challenges: simultaneously optimizing more than ten critical properties before clinical testing. The system addresses binding characteristics including specificity and cross-reactivity, stability factors such as thermostability and aggregation resistance, manufacturability metrics like expression yield, immunogenicity risks including T-cell epitope potential, and developability concerns such as deamidation and oxidation liabilities.

The platform can initiate work from minimal starting points—either a single protein sequence without supporting data or noisy screening pools such as next-generation sequencing enrichment counts from display campaigns. This flexibility allows integration at multiple entry points in discovery workflows.

"Cradle's platform empowers our scientists to explore the design space more efficiently while maintaining full control over candidate selection," said Alistair Henry, UCB's Chief Scientific Officer, according to the announcement.

Scientist-Led AI Integration

Cradle's approach preserves human decision-making authority while accelerating computational exploration. The platform provides both a no-code web application for bench scientists and an API for computational biology teams, enabling different user groups to interact with the technology according to their technical backgrounds.

Scientists receive pre-synthesis scoring of candidate plates, forecasts of expected hit rates, score distributions for each optimization objective, and estimates of remaining optimization potential—decision support tools that inform rather than replace human judgment. Researchers can insert their own rational designs alongside AI-generated candidates on the same experimental plate, with both data streams contributing to model training.

"Our collaboration with UCB demonstrates how AI can accelerate antibody discovery when integrated thoughtfully into existing workflows," said Stef van Grieken, Cradle's CEO and co-founder.

Competitive Positioning in Maturing Market

The UCB deployment positions Cradle alongside competitors including Absci, Generate Biomedicines, and Profluent in the increasingly crowded AI protein engineering sector. The partnership validates Cradle's strategy of integrating into established pharmaceutical discovery processes rather than requiring wholesale workflow redesign—a potential differentiator as the technology moves beyond proof-of-concept collaborations toward routine enterprise use.

For patients, accelerated optimization could compress discovery timelines for diseases with unmet needs, though the announcement provided no specific therapeutic areas or timeline projections. The platform's ability to address developability and manufacturability during early discovery may reduce late-stage failures that delay market access, though clinical impact remains to be demonstrated through actual drug approvals emerging from the technology.

Sources

  1. UCB Adds Cradle's AI Platform to Antibody Discovery Projects — unite.ai