Platelets are small blood particles that play a fundamental role in various physiological processes such as, for instance, the formation of clots that prevent bleeding. Their successful action depends on their adhesion and aggregation capabilities, as well as their transport properties in blood, that is their presence at a vessel wall. Their movement being affected by the shape and deformability of red blood cells within the blood flow, an accurate description of platelets transport is very challenging. Currently, medical devices aimed at the determination of platelets function often fail to detect platelets disorder. Here, we combine discrete numerical models, high performance computing, Bayesian computation with in vitro experiments to infer platelets properties for healthy and unhealthy patients.