By Justin Horne
Learn about sponsoring a Capstone Project
A team of graduate students at the University of Washington’s Global Innovation Exchange (GIX), an institute for emerging technology leaders that’s part of UW’s College of Engineering and Foster School of Business, is tackling a significant challenge in mental health care: how to make mood tracking effortless and actionable.
The students’ project, developed in partnership with experts from T-Mobile and the healthcare community, aims to mitigate the burden of manual mood tracking by using physiological data from wearables and conversational AI. The result is Neuropy: an innovative system designed to provide users with consistent, meaningful insights into their emotional well-being, all without requiring any manual input.
The Problem: Mood Tracking is Essential, But Rarely Consistent
For individuals managing their mental health, mood tracking can be an essential tool for identifying emotional triggers, measuring the effectiveness of treatments, and improving overall well-being. However, current solutions come with major drawbacks.
“People just forget to track their moods, or even abandon it eventually,” said Gwen Zhang, a student in the MS in Technology Innovation (MSTI) program at GIX. “Every single one of the people in our initial survey said that was one of the biggest issues. All of them.”
Beyond inconsistency, existing mood tracking apps often lack deeper insights. “A lot of apps have a bunch of data, but there’s no meaning to it,” Zhang said. “People really want to see analysis and insights.”
This mood tracking is more important than just individual curiosity; for people working through mental health challenges with a medical provider, it’s a key insight to their long-term progress.
“For people with bipolar disorder, for example, it can take two to ten years to find the right medication,” Zhang explained. “That process involves trying dozens of different medications, and clinicians rely on patient-reported data to guide those decisions. If mood tracking isn’t consistent, treatment takes even longer.”
The Solution: AI-Powered Mood Tracking with Wearables & Conversational AI
To address these challenges, the team developed Neuropy, a system designed to provide continuous and effortless mood tracking. The app they’ve created automatically pulls in physiological data from wearable devices, such as heart rate variability, skin temperature, and sleep data. Using established machine learning models, the system estimates mood based on these physiological markers—ensuring a mood entry is always logged, even if users don’t manually report it.
The team also developed a companion hardware device that leverages conversational AI to bring mood tracking closer to journaling. The device, which they dubbed “Home Hub,” allows users to verbally express their thoughts. “People don’t always know how to label their emotions,” said Zhang. “Talking to the home hub is like a brain dump.”
The Home Hub records voice input and analyzes both speech tone and sentiment using large language models. This results in a more natural, passive way to track moods, without relying on the user to select an emotion from a limited menu.
A Smarter Approach to Mental Health Insights
Neuropy doesn’t just track mood; it aims to make the data meaningful. Their app features mood clouds that visually represent users’ most frequently experienced emotions. These data are also tracked over time on mood graphs, which enable users to correlate how they were feeling with other factors in their lives, such as sleep quality, social interactions, and other lifestyle influences.
Lastly, it delivers personalized insights designed to help users understand their emotional trends and triggers, giving them a path towards more direct management of their mental health.
“We put all this analysis into the app so people can explore the mood cloud to understand the connection between their emotions and their behaviors,” said MSTI student Lily Chou. “It gives them real insights.”
T-Mobile’s Role: 5G and Secure Data Handling
The project is backed by T-Mobile, whose 5G capabilities provide critical infrastructure for handling sensitive health data. “In the medical fields, privacy of course very important,” said Zhang. “Intrinsically, 5G is more secure than Wi-Fi.”
Although the team emphasizes that their device is for wellness, and that it’s not a medical device, it can provide valuable insight for them as designers and also for T-Mobile as a provider.
The Home Hub device is also designed with rural communities in mind, where access to mental health professionals is often limited. “One of the use cases that matters to T-Mobile is the lack of psychiatrists and therapists in rural areas,” said Zhang. “Having a device like this in the house would help them engage in tracking their moods and mental well-being consistently. That’s crucial because it’s so hard to see a psychiatrist regularly.”
Hardware + Software for a Comprehensive Experience
The team is leveraging 24/7 data from existing consumer wearable devices with their custom-built Home Hub, which creates a highly comprehensive set of data for their model to evaluate.
“Initially, we looked at devices like Whoop [a smartwatch-like consumer wearable],” said MSTI student Ashwin Nishad. “But we found that they don’t allow you to collect the data. So we moved to the Aura ring and connected to their API which allows us to get that data.”
The team also emphasized the importance of the home hub over just speaking into a phone.
“We want the user to have a great experience. We’ve got a warm light and large button,” said Chou. “We’re trying to give them a very warm experience to feel comfortable talking to the hub.”
“The reason for developing the hardware,” said Nishad, “is also that it gives something for clinicians to prescribe. This device can help fill in the role of staff whose job it is to talk to patients. It’s a super important role, but they’re highly understaffed.”
“Nex, the button creates intentionality,” he added. “Rather than just talking to a device from a distance, this prompts the user to get into the habit of meaningfully making the choice to talk to home hub and engage in the experience.”
What’s Next?
With a fully functional and minimally viable product, the team is now focused on refining their prototypes and preparing for deployment. “The mobile app and Home Hub both work, and they sync to the same database,” said Nishad. “The only thing left is fine-tuning physiological data tracking.”
As their final weeks at GIX approach before they graduate from the MSTI program, the team is focused on polishing the Home Hub hardware and ensuring their mobile app provides a seamless experience. “This is about empowering people to track their moods in a way that works for them,” said Nishad.
By combining AI, automation, and thoughtful hardware design, the GIX team is making mood tracking more accessible, more insightful, and easier to sustain—giving users a powerful tool to better understand their emotional health.
Ready to work on projects like these? Learn more about the MS in Technology Innovation or apply to be a student.
Interested in sponsoring a project? Learn more here.
Back to News
