The transcript of the conversation is fed into a typical large language model (LLM) used by generative AI. But Augmedix has also analyzed more than 500 medical complaints brought by patients to create other data models. They apply both the LLM and the structured data model from their analysis to each transcript in their Augmedix Go service. The use of both modles produces more accuracy and eliminates most oddities (known as "hallucinations" in the AI field). If the two models disagree on an interpretation. the user interface flags the conflict so the doctor can check them.

Learn more about Augmedix: https://augmedix.com/

Health IT community: https://www.healthcareittoday.com/