How AI Helped This Founder Fix a Massive and Dangerous Problem With Prescriptions

Yoona Kim, 44, wanted to help people access health care. That desire led her down a compounding path--from an early career analyzing health care cost trends, to pharmacy school, and then to a PhD program in health economics. Eventually, she encountered an important, solvable problem: Millions of people are prescribed suboptimal--and often dangerous--medications each year. So in 2017, Kim and CTO Penjit "Boom" Moorhead co-founded Arine. The platform, powered by AI, recommends safer, more effective medication regimens. It's grown by illuminating the shadowy side of the $1 trillion U.S. pharmaceutical industry. I always saw gaps in the health care system. My mom is a public health nurse, and even when we were kids, she would bring us to volunteer at public health fairs and county clinics. I remember seeing lines of people out the door--so many people who struggled to get health care. I spent 20 years in health care in different roles. Most of what I did was evaluate the outcomes of different medical interventions. In 2017, everyone was focused on medication adherence and dispensing and delivery solutions for pharmaceuticals. But I realized nobody was focused on the root of the problem: Are patients even taking the right medications in the first place? As a health economist, I'm used to looking at data and evaluating outcomes. My co-founder, Boom, is a nuclear physicist-turned-data scientist and developer. We could've come to market with a manual medication-review service for indi­vid­ual patients. Instead, we spent two years building a platform to be scalable. Humans can't get it right every time. There are so many different data points you need to piece together, like which other medications have been tried and failed, what other conditions the patient has, and their age, gender, and life situation. This is the biggest area of waste in our health care system today--spending $530 billion on the problems that are caused by taking the wrong medicine. We really needed to util­ize AI to solve this massive problem. Before we went to investors, we launched a pilot with a Medicaid plan, and we had some case studies. One of them showed that a patient was taking two to three doses of six separate heart-failure medications. Nobody caught it. He was going in and out of the hospital because of heart failure--probably because of these multiple medications--and every time he went in, he received a new set of medications. Because of the data silos that exist in our system, none of the other doctors could see what other medications were being prescribed to the patient. We have so many stories like this. By making sure a patient's medications are the safest and most effective ones, we can reduce hospitalizations by 40 percent, and that, in turn, reduces the total cost of care. The biggest challenge of turning this platform into a reality was finding our initial seed financing. It was hard to convince investors that changing a medication can positively impact the total cost of care and the total health of an individual--as well as reduce complications and hospitalizations. Nobody knew how big the medication problem was, because everybody thought someone else was taking care of it. I talked to more than 50 investors, and then stopped keeping count. I spent over a year fundraising. It's probably no coincidence that another Asian American woman put in the largest seed check. Oklahoma Health Care Authority was our first customer, and Magellan Health was our first large client. Now, we work with more than 35 health plans and risk-bearing providers, including five ­national health plans, seven Blue Cross Blue Shield plans, and two national pharmacy benefit managers. We've had 100 percent cus­tomer retention since inception. Our biggest driver of growth is focusing on the mission. We need to improve patient outcomes and health for the 20 million lives on our platform. By doing that, the rest will fall into place. More than 50 percent of Americans fill three-plus medications every year. The minimum cost savings we've seen for every patient is $1,500 annually. This is hundreds of millions of dollars for a large health plan. We are just scratching the tip of the surface of the impact we can have.