10 Health Systems and Digital Health Companies to Co-Develop a Safety-Focused LLM for Healthcare
Hippocratic AI has announced the formation of its Founding Partner Program. The program involves leading health systems and digital health companies in both the United States and Canada, including HonorHealth, Cincinnati Children’s, Universal Health Services (UHS), SonderMind, Vital Software, Capsule, and Canada’s ELNA Medical Group, who will play an integral role in developing Hippocratic AI’s technology.
“Across our large network of locations, both nationally and internationally, we recognize that global issues facing healthcare today – such as access to care, staffing challenges, consumer expectations and the need for operational efficiencies – will only grow over time,” said Marc D. Miller, President & CEO of Universal Health Services. “Partnering with Hippocratic AI promises to be game-changing, as we together seek to develop, validate and leverage technologies that will evolve how individuals interact with service providers, and in our industry, transform and improve the way healthcare is delivered.”
Hippocratic AI is building the industry’s first safety-focused Large Language Model (LLM) designed specifically for healthcare, with an initial emphasis of non-diagnostic, patient-facing applications. To build a safer LLM, Hippocratic AI has implemented a multi-faceted approach in creating its product, including: outperforming GPT-4 on over 100 healthcare certifications, training on healthcare specific vocabulary, reinforcement learning with human feedback (RLHF) via healthcare professionals, and working closely with industry experts to verify the model is truly safe.
“We are looking forward to building with Hippocratic AI the tools that will have the biggest impact on improving mental health access, utilization, and outcomes. This work is critical to individuals’ ability to improve their mental health and wellness, as well as empowering clinicians to deliver even higher quality care,” said Mark Frank, CEO and Co-founder, SonderMind
The founding partners will play a pivotal role in ensuring safety and shaping Hippocratic AI. Over the course of their partnerships, they will:
- Guide the development of the model
- Participate in Data Governance and Model Safety committees
- Establish use cases
- Engage their medical professionals in conducting RLHF
- Validate clinical and patient safety
As the teams who currently perform the tasks being considered for Hippocratic AI’s initial use cases, the founding partners are best positioned to determine the model’s readiness for deployment. Each brings expertise, a commitment to safety and innovative spirit, and an eagerness to contribute to the development of Hippocratic AI.
“I’ve been working with AI for years, and understand the importance of models that are trained on specific inputs,” said Aaron Patzer, CEO of Vital Software and founder of the top consumer finance app Mint.com. “There are safe and effective ways to use LLMs like GPT-4 in healthcare today, but ultimately there is a growing need for LLMs that are built, trained, and monitored specifically for healthcare. For us at Vital, partnering with Hippocratic AI means we can more easily build certain functionality for our suite of care experience software in ways that would have otherwise proven to be a challenge.”
“With large language models growing across industries, we are witnessing a technological advancement unlike anything seen in our lifetimes. Bringing this technology to healthcare, in particular, requires an acute understanding of safety and the unique challenges of caring for patients,” said Dr. Jim Whitfill, SVP and Chief Transformation Officer at HonorHealth. “As caregivers ourselves, our team at HonorHealth is thrilled to partner with Hippocratic AI to harness the power of innovation. Together, we will improve the health and well-being of the communities we serve.”
“Our vision is to make healthcare accessible at a scale we’ve never seen before,” said Munjal Shah, Co-Founder and CEO of Hippocratic AI. “Only through the use of LLMs can we get there. These partners are innovative leaders who each have the visions to see that LLMs can radically improve access while deeply understanding that much work needs to be done to ensure safety. It is an honor to work alongside them.”
Hippocratic AI intentionally sought a diversity of thought and expertise for their Founding Program. These partners represent large health systems, patient constituencies, digital health providers, mental health, and children’s health. Some partners have also chosen to invest in Hippocratic AI. As shareholders, they will have additional oversight of the model.
“At General Catalyst, we’re firm believers in the power of radical collaboration to transform healthcare,” said Daryl Tol, Head of Health Assurance Ecosystem at General Catalyst. “Our health assurance ecosystem — comprised of 15 hospital systems and dozens of companies worldwide — merges industry leaders from healthcare, tech, and venture under a shared vision for a more accessible and equitable model of care. This network is deeply interested in the power of LLMs to address staffing shortages and improve care quality, which is why we invested in Hippocratic AI and are thrilled to foster its strategic partnerships supporting their mission to create the safest LLM for healthcare. Each founding partner brings expertise that will be pivotal in training and testing the model. We believe these hospital systems also have the incredible opportunity to be early adopters and shapers of this innovative technology.”
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!
