Munjal Shah, founder and CEO of the pioneering artificial intelligence company Hippocratic AI, has announced an exciting new phase in developing their cutting-edge healthcare large language model (LLM). The company is partnering with over 40 leading providers, health systems, and digital health companies to beta test its LLM-powered products for nondiagnostic services such as chronic care management, post-discharge follow-up, and wellness assessments.
Key partners in this beta testing phase include respected institutions like the University of Vermont Health Network, Memorial Hermann Health System, Fraser Health, and Side Health. By collaborating closely with healthcare professionals on the front lines, Munjal Shah and the Hippocratic AI team aim to rigorously validate their technology’s safety, accuracy, and effectiveness before bringing it to the broader market.
“At Hippocratic AI, we envision a future of health care abundance, where everyone has access to personalized, high-quality care,” said Munjal Shah. “Our partners play an invaluable role in stress-testing our technology and ensuring it meets the highest safety and efficacy standards.”
The Hippocratic AI LLM builds upon recent breakthroughs in generative AI but with a laser focus on healthcare applications. By carefully curating its training data to emphasize evidence-based medical research and insights from healthcare experts, the company aims to create a more specialized and reliable model than general-purpose LLMs like GPT-4. This approach reduces the risk of “hallucinations” or inaccurate outputs.
Early results are promising, with the Hippocratic AI LLM outperforming GPT-4 on over 100 healthcare certifications. The model’s healthcare-specific vocabulary and training via reinforcement learning with human feedback have been critical to this success. Munjal Shah and his team are laying the foundation for generative AI to help alleviate staffing shortages and expand access to care by focusing on essential but lower-risk applications like chronic disease management and post-discharge follow-up.
Beta testing partners are enthusiastic about the technology’s potential. “We believe this technology can support our workforce by automating certain basic tasks, like providing preoperative instructions or post-discharge check-ins,” noted Feby Abraham, Ph.D., executive vice president and chief strategy and innovations officer for Memorial Hermann.
Jessica Moschella, senior VP for high-value care at The University of Vermont Health Network, highlighted the importance of including rural healthcare providers in the development process. “Being involved in this phase allows us to lead the way for rural health care by aiding in the development of new technology that will support our patients’ unique needs and reduce health disparities,” she said.
Throughout the beta testing phase, Hippocratic AI’s LLM will be evaluated by physicians, nurses, and other healthcare professionals on criteria like conversational ability, medical accuracy, empathy, and listening skills. This feedback will be instrumental in refining and optimizing the technology.
With the support of major investors like General Catalyst and Andreessen Horowitz and the guidance of its newly formed advisory councils of health care experts, Munjal Shah and Hippocratic AI are well-positioned to bring their transformative vision to life. As the healthcare industry increasingly recognizes the potential of AI to improve care delivery and patient outcomes, Hippocratic AI’s commitment to safety, collaboration, and innovation sets a powerful example for the field.
