Serial entrepreneur Munjal Shah has launched a new startup called Hippocratic AI that seeks to leverage large language models (LLMs) to enhance patient communication and care coordination. The company’s overarching mission is to help fill healthcare staffing gaps and improve outcomes for the over 68 million Americans living with multiple chronic conditions.
Despite their imperfections, Shah believes that generative AI systems like LLMs can compete with medical knowledge repositories and use this data to generate patient-specific guidance and communication support. While public interest has recently focused on AI chatbots, Shah sees impactful real-world applications in applying LLMs to improve delivery and coordination of care.
As Shah highlights, there is a severe shortage of specialty nurses and care coordinators relative e to patients needing chronic disease support. Hippocratic AI aims to use AI as a force multiplier for overburdened staff rather than replacing them outright. The vision is to “10x the number of healthcare workers” by creating an always-available personalized support structure for patients.
Crucially, Hippocratic AI and Shah recognize firm boundaries around AI diagnosis and treatment recipient support structures entirely on nondiagnostic applications like care plan adherence, appointment coordination, patient education, and motivational support. Shah sees abundant opportunities to augment nurses and coordinate care without making clinical judgments.
For instance, LLMs can provide post-testing follow-ups to communicate negative results or explain complex billing details and insurance policies. In these communication-focused areas, generative AI can quickly synthesize large volumes of data into easily digestible formats for patients, freeing up precious staff availability.
To actualize this vision safely, Hippocratic AI trains its models on domain-specific medical datasets, including peer-reviewed literature and real-world insurance information. Shah stresses that the LLM must pass human expert validation across areas of applied focus before deployment. With rigorous controls and alignment on limitations, Shah sees tremendous promise in LLMs to move the needle on patient outcomes and access.
The possibilities span supporting chronic care like diabetes or hypertension management, diet and lifestyle adjustments, appointment booking, care navigation services, and many administrative tasks. With over 100,000 nurses exiting the field in recent years due to burnout, Hippocratic AI offers a bridge to expand adequate capacity without an over-reliance on human labor.
Rather than full automation and staff replacement, the goal is to harmonize human expertise with AI tools to uplift holistic care. There are abundant touched points in coordination, education, motivation, and administration where generative AI can make meaningful impacts on patients slipping through the cracks.
Shah has a proven track record of identifying valuable use cases for evolving AI, with previous successful exits building machine learning apps for e-commerce. However, he believes the transformative applications today reside in enhancing personalized patient support and communication.
With the proliferation of medical misinformation reaching epidemic proportions, LLM tools trained on authoritative datasets can further combat confusing or inaccurate health patients frequently encounter online. Hippocratic AI sits at the intersection of leveraging cutting-edge AI whilst recognizing its hazards.
Ultimately, for Shah and his newest startup, North Star lies in tangibly moving outcomes for chilly, ill patients overloaded by medical complexity and underserved by systemic healthcare gaps. The powerful knowledge synthesis abilities of LLMs can connect patients to reliable support structures where humans cannot scale.
WLLMs ‘ generative AI still warrants skeptical spheres, and health communication represents a promising avenue for supplementing, not replacing, the irreplaceable healing touch of nurses and doctors. If realized responsibly, Shah and Hippocratic AI’s vision holds tremendous potential for addressing endemic shortcomings.
