Founder Interviews: Artem Petakov of Noom

I started programming when I was nine years old, and I was always interested in everything about artificial intelligence. I worked .com jobs all through high school, and also kept studying the brain (so I could simulate it!) and eventually got more into psychology itself. While studying at Princeton I majored in computer science, but also found myself drawn to developmental psychology books and enrolled in psychology, finance and theater classes (all three of which I consider to be different applications of psychology).

I fell in love with Psychology of Decision Making when I took the class with Daniel Kahneman. Down the road when Saeju (Jeong) and I would go on to found Noom, we wanted to combine those two very things: technology and psychological science.

Before starting Noom, I worked as a Software Engineer and Tech Lead on Google Maps. I started and lead my own project there, and working at Google was an incredible experience (as you can imagine for someone who has been programming since they were nine years old), but ultimately, I knew I wanted to do something where I could have a bigger impact on people’s lives. 

When I met Noom’s co-founder Saeju, I was working at Google, and wasn’t necessarily in search of a new venture—after all, over half of new businesses fail within the first five years, and the cost of failure gets even higher when you launch a new business with a friend. But Saeju waited for me, and had such conviction and belief in the product that we were building, that he convinced me to take the plunge. 

Saeju wrote more about the work we do together last year, but more than anything, our ability to collaborate as partners on this project is a big part of why Noom got off the ground. 

By combining the power of AI, mobile tech, and psychology with the empathy of over 1,000 personal coaches, we help people live healthier lives by changing their long-term habits.

By replacing ingrained unhealthy habits with a healthier way of thinking and behaving, our users are able to actually experience a sustainable change. 

We are starting with weight loss and diabetes prevention, and expanding to other conditions. In the U.S., 72% of the population is either overweight or obese, which is often the gateway into comorbidities. In fact, 69.8% of Noom users reported that they were at risk for one or more other serious health conditions (diabetes, hypertension, etc.).

Looking into the future, the technology that Noom leverages has the potential to help and support those dealing with various chronic care conditions. 

Ultimately, what attracts users to Noom more than anything else is a message that resonates. We all share the goal of being healthy, but what sets Noom apart from other “fad” solutions that users have often already tried out, is that we know results need to be not only attainable but sustainable. Reaching users with that message is ultimately what brings them to us. 

Noom is a subscription-based business that charges monthly fees for the product. We chose our business model incredibly carefully to incentivize ourselves to keep improving the product.

We didn’t want to sell a product that was a one time purchase because we want to be on the long-term journey with our customers. Because we are constantly iterating on the technology and making it smarter through user-driven insights, our business model is what gets improved and optimized every day. 

At Noom, we are doing a lot of things that are new and exciting to engineers at all levels. As a company, our success hinges on our product and engineering capabilities in order to deliver a great experience for users.

As such, our engineers are faced with tough, interesting problems, such as:

Altogether, these are massive opportunities for engineers to truly make an impact and not just be a cog in the wheel. Our engineers work with the latest frameworks that come out (i.e. RxJava, Swift, Kotlin, etc) to build experiences that are around changing people's behavior for the better. 

When it comes to behavior change, it's still unclear what works and what doesn't, and there is a great deal of things to be learned by analyzing the data and building ML-powered algorithms on top of it.

Harnessing the data to understand across hundreds and thousands of users, is the best way we’ll come to understand what is driving actual behavior change. It's a data problem and requires infrastructure as well as clever use of it!