Speaker 1: When Drew Calver, a high school teacher from Texas, suffered a heart attack in 2017, he was rushed to the hospital. He had insurance, yet his hospital bill totaled more than $100,000. That's about twice the amount he was making per year as a teacher. The heart attack was a complete shock to the Calver family. Drew was an avid swimmer, and he had just completed an Ironman triathlon just months before the incident happened. Even though he was in a life-threatening situation, from his hospital bed, Drew asked the doctors whether his insurance is going to cover the cost of his stay. That's a financial concern that accompanies nearly every American who has to go to the hospital. High cost and complexity of care are endemic in our health care system. Health care costs have been rising sharply, at nearly twice the rate of inflation. Fifty-five percent of Americans are worried that they, or someone they know, won't be able to afford the medical care they need. And more than 65 percent of people say they don't really understand how their health insurance coverage works. That's more than half of the U.S. population. In most industries, when we have complex data sets, we can use AI and train models over complex data sets to create clarity. But the challenge in health care is that data is fragmented and distributed all over the place, which means also that it's very difficult to access. And health care costs, ultimately, are hidden until we receive an actual bill in our mailbox after the episode of care happened. Now to the great news. The great news that we don't talk nearly enough about is that the government is now mandating insurance companies and hospitals to release their prices. Federal regulation went into effect about one year ago. And it's mandating that insurance companies and hospitals publish all of their negotiated rates on their websites in machine-readable format on a monthly basis. This data is out. It's been published about a year ago and refreshed every month. The challenge with it is that it's extremely complex and it's sheer size. We're talking petabyte-scale data set, teraflops of information that are published in complex nested JSON structures that are inaccessible to the human eye. And no person on a computer can even open these files. So you need significant compute power in order to build systems to ingest this data at scale and extract insights. When I first learned about this regulation, I was ecstatic. For six years before starting Cascade, I worked at Microsoft on cloud and AI problems, building machine-learning platforms and intelligent cloud systems powered by very complex data sets for optimization and cost reductions to drive efficiency. So I saw an opportunity to finally do that in healthcare. This data is finally available. For the past year of working with healthcare transparency data, we can validate there are wide differences in pricing across cities in America, even for common procedures. Let's take a shoulder MRI. In places like Syracuse or Seattle, prices are more consolidated. But in other places like Boston, Atlanta or Dallas, there is more spread in pricing. And also, there are higher average negotiated rates, which naturally means that people living in these cities have a higher likelihood of receiving a high-cost medical bill. Even when we zoom in on a city, wide variations in pricing remain, driven by the choice of insurance that we each get. What I'm showing you here is something called violin plots. And we are showing distribution of costs across categories of procedures in Boston across the four major U.S. health insurance carriers. Depending on the health insurance network that you choose, you are likely to select care from a category of medical services associated with a pool that's spread out in pricing or that's associated with a more consolidated set of medical rates. Or you're likely to select medical care from sets of procedures associated with higher rates or with lower average ones. It all depends on the insurance you choose, the doctor you see, the medical procedure you have, and where you are located. We have the data to bring transparency to all of this. Until now, it's been hidden. The beauty of our system lies not in the data itself, though, but in making it accessible and actionable for consumers so we can actually use it and make informed decisions before we seek care. We've built the UI for people to ask questions in plain English about medical prices or general health care navigation questions. What's an HMO? What's a PPO? What's the difference between an FSA and an HSA? Which one should I get, given my life situation and background? So you have access to the best informed personal health care assistant. We leverage cloud technologies to transform our data sets on medical price, quality, location, and coverage, along with private client data. And we use large language models to enable users to interact with this data, with the combined data in real time, and optimize decision making. Our platform is a window into how the health care system works so that we drive understanding and enable people to take better decisions. No one should go through having a heart attack and staying on their medical bed and being worried that they or their family won't be able to afford the care they are receiving. We can't prevent medical situations from happening to us. Something will eventually get to all of us. But we can make sure that when something is happening, we are prepared to act with the best available information in critical situations for ourselves and for the case of caregivers in the room, for our kids and aging parents. Today, I'm super excited to announce the launch of our conversational AI agent powered by health care transparency to the first 500 people to sign up in Boston. So you can query prices for medical procedures for office visits and for emergency room cash prices. You can scan the QR link and sign up for this live. If you're an employer in the room, please reach out. We can also enable this solution for your entire workforce and personalized for your demographic and personalized for the specific network that you are offering. Thank you so much, Boston. It's been a pleasure.