Hippocratic AI: Revolutionizing Healthcare with Generative AI

In this episode of Friends of Project Healthcare…

Vice President and Head of Partnerships at Hippocratic AI, Nick Archer, joined host Eric Thrailkill and Project Healthcare colleague Lauren Hatcher, to discuss some of the challenges with information asymmetry in healthcare, and the negative impact it can have on patient outcomes. 

Hippocratic AI is working to solve that burden by building the first safety-focused large language model (LLM) for healthcare to reduce workflow misalignment with conversational AI.

Hippocratic AI launched on May 16th, 2023 with over $50MM in raised capital from a A16z, General Catalyst company investment and its healthcare founding partners which include a group of physicians, hospital administrators, and AI researchers from Johns Hopkins, Washington University, Stanford, University of Pennsylvania, Google, and Nvidia.

Since the advent of Open AI’s Chat GPT, there has been a tremendous interest, particularly in the healthcare space, for generative AI functionality created by the rapid acceleration of LLMs.. 

Archer and his team are eager to leverage the opportunities offered by gen AI to revolutionize the approach to care, specifically by reducing workflow burden and increasing efficiency. 

Some of the many use cases for their platform include assisting with post-discharge follow-ups, pre-op assessments and other administrative tasks, all of which fall under a concept he calls “super staffing.”

“When we’re trying to brainstorm with potential partners around use cases, we tell them to look at it from a super staffing perspective. What are the things that you’ve never considered doing because you could never hire enough staff, or it just wasn’t affordable, or maybe time was the enemy?” Archer posed. 

But Archer was quick to point out that despite the significant promise, safety is paramount, and highlighted a use case Hippocratic AI will not delve into anytime soon.

“We’re staying away from diagnosis,” he said. “It’s a question that we get asked a lot moving into meetings, and we’re not enabling it. We don’t think it’s safe yet.”

As its namesake implies, the same oath medical professionals take to “Do No Harm,” Hippocratic AI recognizes the need for a safety-first approach, and Archer pointed out that regulatory uncertainty and accuracy risk with LLMs built from public data sources is of great concern. 

Archer argued a safe model needs to understand medical and clinical language, and the association between different terms. It also needs to comprehend drug names and potential negative interactions. 

“One of the keys and what’s really important about having a vertical model in healthcare is safety. And so a language model, its comprehension, is based on the data that’s inside of its brain,” Archer explained. “And so we have filled our brain, our LLM with a tremendous amount of healthcare information.”

To further address the safety issue, Archer shared some of the “secret sauce” behind their LLM. Unlike other models on the market that have used content from different public sources, Archer revealed all of the content fed into his LLM is purposefully acquired or licensed from trusted resources.

“We are getting data that’s not widely found, specifically healthcare data, to really improve the accuracy and the comprehension and the use of our model in healthcare settings,” he said. “We’re building for safety-first and that’s the focus of our go-to-market.”

Building trust with the patient and medical teams alike is crucial to the adoption of AI in healthcare. Hippocratic AI prioritizes patient safety and its accuracy by incorporating reinforcement learning with human feedback (or RLHF) using healthcare professionals. 

“We’re doing our RLHF with clinical professionals,” he revealed. “We don’t want to ship a product that a clinician says is not ready for primetime. So we are actually hiring nurses and we’re doing the RLHF on these tasks that we’ve been discussing with clinical professionals. That’s a component of safety for us.”

To make good on their goal for inclusivity and improved accuracy, Hippocratic AI partners with health systems, digital health organizations, payers, and even pharmaceutical companies to leverage the power of their LLM.

“We’re working with companies around the world who have an interest in leveraging our product and platform,” Archer shared. “Some of what we work through with our product team and with our partners trying to understand the nuance of their use case and how we can best use the language model and these voice-based capabilities within this specific problem they’re trying to solve to drive value.”

By collaborating with these partners, Hippocratic AI aims to drive innovation, improve workflow, and enhance overall patient care. “This shouldn’t be something that happens to clinical professionals. This is something that we want to do with clinical professionals and with health systems,” Archer explained. 

“It’s one of the reasons why we didn’t just build a product and start shipping it out in the market. Literally, we are building these use cases right now with our partners. We’re going to go to market with them and pilot with them in the same manner. We want clinical professionals to play a role in what we’re building and in what we launch.”

About Nick Archer: Nicholas (Nick) Archer is the VP of Partnerships at Hippocratic AI. Hippocratic AI is building the first safety-focused large language model (LLM) for healthcare. The company has raised a $65MM seed round from A16z, General Catalyst and its healthcare founding partners. Before joining Hippocratic AI, he served as an executive for 19 years at AdventHealth, a $15B national health system. During his tenure, he held executive roles spanning finance, strategy, and product innovation. Passionate about giving back, Nick has been actively involved in numerous charities, serving on the boards of United Against Poverty, Andrews Memorial Hospital in Kingston, Jamaica, and Junior Achievement.

He holds a bachelor’s degree in finance from Oakwood University, an MBA from the University of Florida’s Hough Graduate School of Business and has completed Executive Education from Harvard Business School’s Launching New Ventures Program. In his spare time, he enjoys time with his sons, cycling and CrossFit.

Keep the Pulse, Friends!

LISTEN: “Follow” or “subscribe” to Friends of Project Healthcare on your favorite podcast platform.

REVIEW: Go to Apple Podcast at projecthealthcare.com/apple and leave a (hopefully) 5-star rating and creative review.

Friends of Project Healthcare Podcast

Brought to you by Nashville Entrepreneur Center