A Nature Health study introduced DocTr, an AI system that recommends investigators and sites for clinical trials by analyzing patient encounter data, unstructured protocol documents, and historical enrollment records. The authors reported DocTr achieved 58% higher similarity in clinician recommendations versus best‑in‑class baselines on unseen trials and showed robustness across phases and disease areas. DocTr operates upstream of patient‑matching tools by optimizing site and investigator selection to improve recruitment probability and outreach efficiency. The technology is pitched as complementary to patient‑level matching and as a way to reduce site activation risk and improve enrollment timelines. Sponsors and CROs facing persistent recruitment bottlenecks may evaluate DocTr to refine site selection strategies and reduce operational costs associated with underperforming investigators.