Putting AI to work in helping patients with chronic illness, juli raises $3.8M

The company is focusing on asthma, migraine, depression, bipolar disorder, and chronic pain

There's an influx of data in the healthcare system, with a single patient generating as much as 80 megabytes each year in imaging and EMR data. At the same time, chronic disease is rampant in the U.S.; according to the CDC, six in 10 adults have at least one chronic disease, and four in 10 have more than one.

Yet, despite both of those trends, that data still isn't being used to help patients who have chronic illnesses, says Bettina Hein, CEO and co-founder of juli, a platform that uses AI to help patients and their care teams manage complex chronic conditions. On Thursday, the company announced a $3.8 million seed funding round led by Speedinvest and Norrsken VC, with participation from SoftBank’s Vision Fund Emerge Program  and Dieter Von Holtzbrinck Ventures.

Hein encountered that disconnect herself when she fell while pregnant, causing her daughter to be born premature. That, in turn, caused her to start losing sleep, something that her doctor couldn't help with.

"It turned out that I had an atypical postpartum depression. It would seem logical for a PCP to screen for that, but it was atypical and I probably wouldn't have shown up too high there, anyway," she told me

"When I cycled off my tour of duty as CEO of my second company after 10 years, I decided it was time to use my skills in an area that needed them. My background is in taking these large data sets and applying machine learning to them, but why wasn't that being done for chronic conditions?"

What juli, which stands for "just use lots of information," does is take all of that data, be it patient-provided, environmental, and even the patient's social and behavioral context, and use it to identify ways to improve their health.

"For example, environmental data plays a huge role in depression; we found out from our data that the third highest correlating factor for depression is air pollution. So, if we see that, it's really interesting what you can give people for advice," Hein explained.

"You could say, 'get an air filter,' or you could say, 'go jogging after a rainstorm because the air is cleaner, there's less pollen is in it.' That's just one example and there's so many of these things that are undiscovered."

For patient who use juli, they just have to download the app and then go through the onboarding sequence where they pick from the conditions that juli currently supports, which include asthma, migraine, depression, bipolar disorder, and chronic pain; it will soon be adding long COVID in as well. The patient tells juli about their own health, including if they've been diagnosed, if they see a doctor, and if they take medications. Patients are then asked to self-report daily about how they're doing by picking one of five smiley faces.

The company then gathers that data and puts it together to assess how the patient is doing, using artificial intelligence to predict where they are on the standardized scales for their specific conditions, such as the PHQ-8 for depression, the Asthma Control Test for asthma, and the Headache Impact Test for the migraine sufferers.

"We take all the data that we have around your health and we look for triggers and for levers. This is the journey that most people with chronic conditions are on: they want to find out, pretty desperately for a lot of them, what's wrong with them. What is doing this? If you have a complex condition, there are so many different factors that it's hard for the human brain to do that and the solution to this point has been all your different doctors tell you to have a diary, on paper oftentimes, but none of that fits together," said Hein.

"So, with AI, it's actually pretty simple to pull this stuff together and then take it apart again, to figure out what the correlations are. That's cool because people become more aware of it but also we can do predictions."

For the doctor, it helps them know which patients need help at the moment, which can lead to better outcomes for the patient: launched less than a year ago, juli already has 15,000 patients who have signed up and it has already been able to show that it can lower the PHQ-8 levels for those who have depression by 22% in eight weeks, on average. The company has also been able to cut in half the number of people that have poorly controlled asthma, and "move people down significantly" on the pain scale for migraine.

The company plans to use the new funding to grow its team, specifically around marketing and data science machine learning, as well as to conduct clinical trials of the app, which Hein says will help the company prove its viability in a crowded marketplace. It will also continue to build out it's AI capabilities.

"We have so many variables that we are collecting, figuring all that out and training models to then actively give people advice and to get them to make these micro behavioral changes that they need to make to get better," said Hein.

"The more data we have, the better we get to know a person, the more powerful that becomes because the people that we have that that are our patients see that, you have to invest in this in order to get something back, we have a lot better retention rates then other apps out there because it gets more valuable over time." 

In the long run, the goal for juli is to use that data and those capabilites to get patients the care they need quicker.

"If we can just shorten the journey to stability for someone with a chronic illness by six months then we will already have accomplished a lot. It's about making people better."