[GAME CHANGER]

People crave a human touch, even when they use digital health care platforms.

Sure, working out at home using a fitness video has its perks, like saving time and money. But guidance from an expert — especially in the beginning, when correcting posture or getting recommendations for physical therapy exercises — is an option people like to have, especially to get rid of that persistent feeling that something isn’t going right.

Meshing the best of both worlds are digital health care platform firms like iPixel. Consult a physician one-on-one to get the basics right, then work out on your own time in the comfort of your home using a posture-correcting app.

“Think back on the times you had a personal trainer. They put in a lot of effort giving out instructions at first but as the sessions go on, they’re just counting reps while you go through the same motions — it feels like a waste of money,” iPixel founder and CEO Lee Sang-soo said.

The company has two products on the market. The first is Exercite, an exercise app directly sourced to users; the second is Exercite Care, specializing more in digital musculoskeletal (MSK) therapy. Both platforms use AI-powered computer vision technology that tracks the user’s motions and provides real-time feedback.

Exercite Care has a few extra steps, starting with an appointment with a licensed doctor or physical therapist who recommends exercises to the patient. Data is collected while using the app and is sent back to the expert to analyze and give feedback at the next meetup. Rinse and repeat for a few months.

The Korean startup, however, is relying on more than its computer vision technology to get ahead. It competes with global players like Hinge Health, which links up with a wearable device (Excercite doesn’t), Kaia Health, Omada Health and Sword Health as well as domestic rival EverEx.

It’s also looking for deals with hospitals and senior citizen homes. Doctors and therapists make money by providing consultations — which iPixel says it does through a quick and easy user interface to assign exercises and provide feedback — while the software company charges customers to use Exercite Care. While the business-to-business model is currently in place for Exercite Care, iPixel wants to expand it to Exercite and link it up with gym coaches in the future.

There’s also a local advantage to launching in Korea — the startup hopes to piggyback on the country’s electronics giants to deliver its services to more customers. Exercite is currently available on LG Electronics’ smart TV platform webOS.

The firm’s future goals now hinge a lot on whether Exercite Care will be approved for digital therapeutics (DTx) licensing, which would open the door for iPixel to be prescribed as medical intervention by doctors — a big first step to prove itself in the domestic market. The software is currently being provided as a service at Incheon Sejong Hospital and Bucheon Sejong Hospital, but doctors have been using it primarily for research purposes and are not paid for their consultations.

iPixel CEO Lee shared the startup’s journey so far and its ambitions to expand overseas in an interview with the Korea JoongAng Daily.

The interview has been edited for length and clarity.

Q. Your background is mainly in game development. What motivated you to enter the digital health care sector?

A. I’ve always loved tech, but I was especially fascinated by augmented reality. I was especially intrigued with Pokemon Go when it came out in the summer of 2016, and knew I had to create some sort of digital content that allowed users to interact with their surrounding environment.

Then I got to learn about motion recognition, and initially developed the software with users who want to train at home in mind.

It was the hospitals who contacted us first after seeing our product. They wanted a program that could be used for physical therapy. Exercite Care was borne out of market demand, and we’ve been focusing on vertical growth in the wellness sector ever since.


What makes your product different from competitors? What does the company’s patented technology do?

Our platform can be thought of as YouTube dressed up with AI. All hospitals possess, and want to recommend to patients, their self-produced videos of physical therapists demonstrating exercises. We use data on some 2,300 motions to label and categorize the exercises in those videos and connect it to our system. Experts use this categorized data to assign exercises to patients.

The second is that our platform is accessible on televisions in addition to mobile phones. Smartphones can be difficult to use for older people or even for younger people who don’t want to use small screens. It can be difficult to set up the camera to fit your body into the frame. Televisions have optimally large screens but have poor processors. Our software optimizes AI for smart TVs, and is the only one that can link up with the devices, even compared to U.S. competitors. That’s one of the patents we have.

When we developed the software, we wanted to make it accessible without having to pay for expensive equipment. The software is therefore available for widespread use across all hospitals and user devices.

How has the product been received so far?

We recently did a study with a major electronics company, where we invited people in their 70s to use our app. A physiotherapy professor recommended exercises to them.

Eighty percent used our app for three consecutive weeks. That’s a great number in terms of retention. The company has already asked to try a second trial for proof of concept.

What it means for mobile phone or television manufacturers is that including our service when selling the devices to senior citizens would be effective.


Where was the data used in the software sourced from? Is there a measure to prove its motion detection accuracy?

We sourced our own data. It was a lot of grinding. Staff members went through the motions while looking at training videos and made recordings that would then be sent to fitness experts or physical therapists to supervise; that data will then be fed into machine learning. When we first developed the service in 2021, we had around 2,300 units of blockchain fitness data.

We also organized the experts’ opinions on which muscles had to be engaged, and which motions were important to constitute good posture, and inserted that into the mapping; for instance, having to use core strength for squats.

It’s obviously more accurate in a lab environment, almost 100 percent, but in an ordinary setting, the accuracy is at around 95 percent. It’s more accurate than what a physical therapist could determine by sight.


How far has iPixel’s technology progressed in the DTx licensing process?

We have been approved for good manufacturing practice (GMP). We’ve now signed up with the Ministry of Food and Drug Safety to hold clinical trials. For medical devices like ours, the trial only needs to be held once. The results typically come out around three months later, but in the current situation, we’re thinking six months due to the doctor walkout. I think early 2026, March at the latest, is a reasonable timeline to have it finally licensed.

Big digital MSK players in the U.S. market are partnering with clients like the U.S. Air Force, Amazon and Uber. What companies are iPixel in discussion with?

We partnered with SK hynix before for use among factory workers but currently don’t have such clients. It’s a different business environment in Korea than in the United States because of the different national insurance systems. U.S. firms pay for their employees’ insurance, and so they save money by keeping their workers healthy.

We’re looking for partners in senior citizen homes and communities. But I also think the Korean medical system will change to adopt systems like ours as they are efficient.


Hinge Health’s recent initial public offering (IPO) filing has garnered a lot of attention. Reports suggest that it indicates a growing momentum for digital health care, which stagnated after the Covid-19 pandemic. Why do you think investors should be interested in the sector at this moment?

I actually think the demand for digital health care has been constant.

It’s true that traffic for digital health care was especially concentrated during the pandemic, but it’s more like demand exploded at an abnormal rate then, and has since grown steadily at a more normal rate.

I’ll give you an example: Indoor bike and fitness content firm Peloton’s value skyrocketed in the United States in 2021. Its stocks then fell in 2022 as hardware sales were stunted, so it seemed like the market had shrunk.

For software companies like us, or in the case of Hinge Health, the need for digital content is constant. And as studies start to show that the apps are effective in MSK care, and the firms who pay for the users’ insurance fees feel the difference, it accelerates the company’s growth.

What are iPixel’s overseas expansion plans?

We’re already seeing some traffic from the United States since our service is on LG devices.

We’ll need a local partner in the overseas market, and it’ll be easier since we’re a software company. We’re not looking to launch an overseas unit. We’re in discussions with some interested investors from the United States, maybe for Japan.

Since the consultation stage is an important factor for Exercite Care, we’ve been visiting events such as those held by the American Physical Therapy Association and the American Medical Informatics Association to look for potential partners.


What are your future investment and IPO plans?

We plan to secure investment once more to accelerate the DTx licensing process and expand our market, especially as it is highly likely that profitability will grow after its approval.

We’re aiming for early 2028 for an IPO.

Software companies’ expenditures are mostly fixed costs, and while it’s hard to cross the break-even point, it’s all profit from there. iPixel only needs to cross 1.5 billion won ($1 million) to make a profit.