The BlueSignum team at Google for Startups Accelerator.

In 2019, Seoul National University students Jeonghyun Yoon and Yoo-Jin Jo created a social robot that could provide emotional support and companionship for people living alone. They built the physical robot and software based on the Robot Operating System (ROS), a set of open source software frameworks for robotics development. Yoon and Jo developed an Artificial Intelligence (AI) engine to analyze the moods and emotions of users when they spoke to the robot and worked to program the robot to react appropriately during specific interactions. “We became confident that software could have emotional value, and that tech could break down barriers for people living alone,” says Yoon, now the CEO of BlueSignum. “When we were working on creating robots, we realized they sometimes can do more than people.”

Jeonghyun Yoon and Yoo-Jin Jo’s college AI robot.

Because many people who struggle with mental health challenges face transportation and cost barriers to accessing a doctor or counselor, Yoon and Jo decided to build a mental health tech solution that people could easily access via their mobile devices to provide positive emotional support. BlueSignum launched with a simple mood tracker app, Daily Bean, which allows users to record their feelings and self-care actions—such as nutrition, exercise, work, and social events—throughout each day. They can track their progress and patterns over the course of a week, month, or year. BlueSignum uses Firebase to analyze the app’s user data and improve customer service.

Now, in collaboration with a research team from the Department of Psychology at Seoul National University, BlueSignum is developing a digital mental health guide that combines cognitive behavioral therapy techniques and Artificial Intelligence to help people self-manage their mental health.

Banking on habit app potential

Investors have categorized Daily Bean as a “habit app” or “routine app” that helps users develop healthy habits. According to Kurdo.org , habit apps often experience quick user growth and are focused on digital therapeutics and chronic disease management, making them attractive to investors. “The essence of the habit app is to induce behavior change,” said Yoon-seop Choi, CEO of Digital Healthcare Partners, BlueSignum’s first incubator and seed investor.

Blue Signum aims to go beyond the habit app space and build out its more complex services, establishing itself in the global healthcare app market. The AI Listener service BlueSignum is currently building will allow users to have a conversation with an AI speaker. The app will use speech to text (STT) voice recognition to capture nuances in users’ voices and natural language processing (NLP) to analyze users’ language when they’re talking about how they feel. It will use machine learning to classify the user’s psychological type and provide mental health guidance.

The BlueSignum executive team.

Google for Startups Accelerator and Google mentors help BlueSignum develop a new service

Once BlueSignum secured seed funding, Yoon and Jo wanted to get insight into the global market and technical mentoring from senior engineers. Their seed investors suggested they apply to Google for Startups Accelerator: Korea.

BlueSignum’s Google tech mentor, Stefano (Woohyuk) Jang, is a Google Cloud data analytics specialist , and their Startup Success Manager Jerry Jeong was a partner engineer on the Google Cloud team. Yoon, Jo, Jang and Jong met weekly, with Jong and John commuting an hour to BlueSignum’s office each week. “Our mentors understood our business model and services and helped us develop our core engine,” says Yoon. “We still contact them every week and ask them questions.”

Jang and Jeong gave BlueSignum advice on which indicators should be measured first in startups, how to collect logs, and what procedures are effective when creating a machine learning model. They talked with the founders about their business model and backend architecture for their AI Listener service. They advised the team on data collection and Artificial Intelligence/Machine Learning systems (MLOps) using Google Cloud and Google's AI tools, and they helped the engineering team build an MLOps environment based on Vertex AI. The mentors also encouraged BlueSignum to use Cloud Run-based backend service architecture in their new AI Listener service.

Blue Signum ultimately aims to leverage a range of health care data, including voice recognition, workout data, and heart rate data, to provide psychological therapy using a highly personalized psychotherapy AI model. To work toward this goal, they are developing a model in a Google Cloud Platform Vertex AI environment, using Google's speech recognition and natural language processing tools.

“We use Google Cloud technology to increase efficiency and speed and implement all the AI engines we need,” says Yoon. “It’s important for us to move quickly and improve our user experience quickly, and high-quality Google tools are of great help.”

BlueSignum co-founder and CEO Jeonghyun Yoon.

What’s next for BlueSignum

Yoon says since BlueSignum’s services will take time to build and create value, the team is waiting to establish product-market fit before pursuing more investment dollars.

“We’re excited to launch more services this year and hopeful that our mental care services can actually help people who need it,” Yoon says.

Yoon says her team is the best part about working on BlueSignum.

“I have the best colleagues,” she says. “Every team member is competent and has excellent teamwork skills. They pursue all tasks efficiently and communicate very well.”

Yoon says she’s experienced great personal growth as an entrepreneur, overcoming moments of feeling limited because she didn’t have much industry experience.

“Like me, I hope other founders do something they truly love,” she says. “Affection for the business gives founders the power to overcome many difficulties. It is not an easy journey; there are many challenges. The biggest lesson is that all the hard work pays off in the end.”