OpenEvidence's approach to training AI models is distinct; it does not connect to the public internet, ensuring that the information provided is based solely on vetted medical research. Additionally, OpenEvidence has formed a strategic partnership with the New England Journal of Medicine, which is unique as they typically do not allow AI companies to train on their research.

OpenEvidence is trained on peer-reviewed medical literature, serving as an AI co-pilot that assists doctors in making better decisions at the point of care, which is undeniably beneficial for humanity.

The use of OpenEvidence has already saved lives by helping doctors make better clinical decisions based on accurate medical information.

By making OpenEvidence freely available, we enabled widespread adoption among U.S. doctors, with 10-20% starting to use it overnight to access new medical studies and improve diagnoses.

Open Evidence has the opportunity to aggregate medical knowledge from millions of physicians worldwide, encoding and improving care by leveraging distributed clinical wisdom.

Open Evidence is a medical search platform that has grown to support over 25% of US practicing physicians as monthly active users, with many using it daily for critical information.

Our approach to AI in medicine involves using smaller, specialized models trained on peer-reviewed literature that outperform larger models on specific medical tasks, which was a novel insight at the time of development.

The concept of 'open' in OpenEvidence signifies a direct approach to doctors, addressing their pain points without gatekeepers, and making medical information more accessible.

The editorial board members of the New England Journal of Medicine were already power users of OpenEvidence, which led to their willingness to collaborate with us.