Unlearn.ai builds digital twins of clinical trial participants, forecasting how each would have progressed, and uses those forecasts to reduce how many people the control arm needs. Scout searches literature and regulatory material, Hindsight explores datasets, SimLab models trial scenarios, and monitoring runs during the trial itself.
A 33 percent smaller control arm is the number that matters and it is an ethical argument as much as a commercial one. Every control participant receives a placebo or standard care while a candidate treatment is being tested, and needing fewer of them means fewer people in that position and faster enrolment, which the company puts at four-plus months saved.
It works with pharmaceutical and biotech companies in neuroscience, immunology and metabolic disease, citing up to 20 percent sample size reduction and $250K of per-patient programme savings. No pricing is published. Digital twin methods need regulatory acceptance for each specific trial context, a forecast substituting for an observation is a statistical argument reviewers must accept, and the savings figures are the company’s own.
