With Monolog, Journaling Meets Artificial Intelligence

Leigh K. Sharpless ’23 and Jerry S. Yang ’23 believe your phone can help you understand how you’re feeling. Monolog, the journaling app they created during their pandemic gap years, is designed to do just that.

With Monolog, a user can describe their day through text or speech. After a few logs, the app will notice trends across their entries and identify how they’re feeling about certain events, places, or people. One post on Monolog’s Instagram page reads: “This month, your discussion of Mom is correlated with higher anxiety than it was last month.”

Monolog tries to differentiate itself from other mental health apps by using natural language processing, a technology which allows devices to understand text and spoken words like humans do. According to the app’s sleek, purple website, Monolog is meant to be quicker than traditional pen-on-paper journaling and more accessible than therapy. Its tagline reads: “Understand your Emotions. Understand your Trends. Understand your Story.”

The idea for Monolog came out of observing trends in declining mental health during the pandemic. “From celebrities like Simone Biles dropping out of the Olympics due to mental health reasons, all the way to physicians and providers calling this a national state of emergency related to mental health, we were seeing more and more people caring about this,” Sharpless says.

Sharpless and Yang found that this trend also applied at Harvard, particularly last fall, when they observed that morale among students living on campus was low.

“When we came back, we noticed that there was this general sentiment of lethargy and confusion and lack of enthusiasm,” Sharpless says. “It was quite evident that Covid had taken its toll on people.”

Seeing a need for a tool to manage mental health struggles, Sharpless and Yang decided to take the year off to develop Monolog. Using Yang’s background in natural language processing, the co-founders were confident that they could approach tracking mental health differently.

Monolog’s technology is designed to quantify and draw conclusions from the way users talk about particular subjects.

"We’ve trained a model to pick out topics specific to what we’ve seen in journaling a lot,” Yang says. “So you can see this trend in terms of how you talk about a specific person and what tone you’re using over time.”

The model also identifies how positively or negatively the user discusses the topic. Over time, users should be able to identify trends in their behavior that might not otherwise have been clear.

“It felt like a pretty natural way that we could add value on top of conventional journaling,” Yang says.

Over the next few months, Yang will be continuing with Monolog and Sharpless will be phasing out of the startup. Right now the app is operating in a closed beta; some of the functionality Yang and Sharpless describe is still in development.

"I think we’re not where we could be in terms of products that actually are effective," Yang admits, but he’s optimistic for the future, planning to add features in conjunction with user feedback.

In the long run, the team hopes Monolog will become an “all-encompassing wellness platform,” Sharpless says. “The place you go for personalized journaling, for personalized content recommendations, or maybe articles about specific topics that you’ve talked about in your journaling.”

Sharpless describes the app as an essential compromise between intensive and lax strategies to address mental health challenges. She adds, “We’re trying to make that middle ground, more convenient, more handheld approach to helping people understand themselves.”