Ambience Healthcare has introduced a new artificial intelligence model designed to code medical data with greater accuracy than physicians. Built using OpenAI’s reinforcement fine-tuning methods, the model focuses on ICD-10 coding, a complex system used to classify diseases and health conditions. In internal testing, the model showed a 27% relative improvement over physician benchmarks. To validate the results, the company compared the AI’s performance against a panel of 18 board-certified doctors using expert-agreed clinical scenarios. Ambience said this advancement could significantly reduce billing errors and improve administrative efficiency in health care.
Brendan Fortuner, Ambience’s Head of Engineering, told CNBC, “We’re not replacing doctors or coders. What we’re doing is we’re liberating them from administration, and we’re fixing mistakes that help make health care better, safer, more cost-effective.” The company’s technology records patient visits with consent and generates clinical documentation in real time. Its tools are currently used by over 40 healthcare organizations, including the Cleveland Clinic and UCSF Health. Ambience is also exploring similar applications in areas like CPT coding, prior authorization, and clinical trial matching. Backed by investors such as Kleiner Perkins, Andreessen Horowitz, and the OpenAI Startup Fund, the company has raised more than $100 million.
