Olio Labs AI Platform Predicts Human Clinical Trial Outcomes

The research, published August 5, reveals that standard rodent models often fail to translate to human outcomes because they rely on narrow, biased endpoints. Olio Labs replaces this fragmented approach with AI vision models trained on expert-labeled data. These systems capture subtle behavioral shifts in animals following drug administration, feeding that information into algorithms calibrated against human clinical trial records.

In a recent study, the platform successfully predicted gastrointestinal side effects and neuropsychiatric toxicity within a 24-hour window. Notably, the system outperformed traditional two-to-three-week studies in predicting human weight loss, achieving 70% higher accuracy while operating 20 times faster. CEO David Tingley noted that as drug discovery accelerates, developers face a critical need for better tools to prioritize candidates before they enter expensive human trials.

CTO Tom Roseberry argues the data discrepancy in historical studies was never a failure of animal models themselves, but a failure of measurement. The company is now applying this technology to develop its own combination therapies and is opening its platform to industry partners seeking to refine their preclinical pipelines.