OpenEvidence

OpenEvidence is the physician-facing medical AI product discussed in E227｜美国医疗市场AI争夺战：巨头押注，创业公司能赢吗？. The episode describes it as a tool for doctors that answers from top medical journals, authoritative clinical guidelines, and licensed sources rather than relying only on a general model’s latent knowledge.

Its wiki role is to make Evidence-Grounded Medical RAG concrete. 周叶冰 / Zhou Yebing treats the product as a response to ordinary AI Hallucination and weak citation quality in medical answers, while 张璐 / Zhang Lu frames its moat around content authorization, physician adoption, and workflow placement rather than only base-model capability.

Key Points

• The episode says OpenEvidence’s valuation reached about $12 billion and annual revenue was around $100 million.
• Doctors are described as using it for professional knowledge retrieval, with free physician access and revenue coming mainly from medical advertising, content promotion, and possible enterprise versions.
• The product’s strongest safety claim is grounded answering with sources, but its strongest business-model concern is whether pharma promotion can bias what doctors see.
• OpenEvidence links the wiki’s earlier Medical Literature Search branch to current Retrieval-Augmented Generation and Medical AI Workflow Integration.

Connections

• Evidence-Grounded Medical RAG — main concept the company makes concrete.
• Medical Literature Search, Retrieval-Augmented Generation, AI Verification, and AI Hallucination — knowledge-access and reliability context.
• Medical AI Marketing Risk and Medical Platform Trust Crisis — trust risk if advertising changes medical answer ranking or display.
• 张璐 / Zhang Lu and 周叶冰 / Zhou Yebing — guests interpreting the product’s market and clinical value.