Andrea Campos, 26, is the founder of You Are Not Alone, a Spanish-language artificial intelligence (AI) mental health app in Mexico. She is a self-taught programmer and a graduate of Dev.f, a coding school based in Mexico City, where she still resides.

A very personal problem

Around the time I started getting treatment for my lifelong depression, I was learning to code. When I started cognitive behavioral therapy, the first thing I thought was, “Oh, my God, I could automate this.” 

When I built Yana, it was not initially meant to be a startup. It was, for me … I wanted to build something that could help me survive my next depressive episode. 

In Mexico, very few people seek some form of help for mental health. People may think that, if you go to a therapist, you’re crazy. And therapy is not covered by health insurance, so you still have to pay out of pocket, and it’s very expensive. 

Out of the entire budget that Mexico has for health, only 2% is used for mental health; 90% of that 2% is used for psychiatric hospitals. 

Owning yourself

We built an earlier version on Facebook Messenger, because Facebook is cheaper for people, as social networks are included in data plans. At the time, we were really confident launching on Facebook, because people wouldn’t have to download another app on their phone. It wouldn’t take up extra space on their phones. It was a platform they were already familiar with. We didn’t have to develop a new UI. 

But we were having difficulty maintaining it, because Facebook Messenger policies change very quickly. If Facebook were to have a major outage, we would lose all our contacts. We realized that relying on a platform you don’t have control over is stressful.

We also found that not having a standalone app was confusing for people. Young people don’t really use Facebook, and some were confused why they had to download Facebook to use Yana. When we went to meetings with schools to pitch Yana as a product for their students to use, they’d say, “Where’s the app?” even when we explained it was built on Messenger.

Hiring developers is a challenge in Mexico. There are still not many developers here, so they’re in high demand and charge American prices, even though they may not have much depth of knowledge. I tried hiring an in-house developer and even hiring a team in San Francisco. None of it really worked. We needed someone who cared about the product.

After running into roadblocks with Facebook Messenger, we made the pivot to creating our own Android app.

I was obsessing a bit about building Yana in Python, because I wanted it to be able to scale. Python is a much more reliable language. Python is also commonly used in AI, data science, and scientific projects, while PHP is more ideal for web development. But it was easier for them to build in PHP, and we needed to launch first. 

Finding the right person, not the best

The most important thing I learned was how hard it is to work with a developer you aren’t comfortable with. You need more than someone who can just do the work; the thing that matters is picking the person that wants to solve the problem.

I hired a new developer. He didn’t have a lot of coding experience, but he could solve problems, and that was more important, so I taught him to code. He brought on another developer from Bolivia, and we had the Android mobile app done in a month. 

We launched at the beginning of the pandemic in Mexico. 

Right now we have 6,100 users, and we have at minimum 400 using the app per day. A lot of them are worried about the pandemic, but it’s different in different age groups. Users from 13 to 17 are mostly affected by boredom or frustration; people from 18 to 30 are mostly affected by uncertainty about the future; 31 to 60-plus fear getting infected.

On one level, the pandemic has helped us a lot — and I hate to say that, because it feels very selfish — but people are in their homes, in this naturally stressful environment. Many people are using Yana now, and much of that has been through word of mouth. 

On the other hand, it’s made the launch of our iOS app much harder. We decided to pause our marketing efforts until we launched on iOS, because we didn’t want to lose those users by advertising something they couldn’t use.