AIMHealth Innovation Lab at Florida State University

GLOW

GLOW

A web-based AI chatbot designed to tackle behavioral change via DBT and chain analysis.

OVERVIEW

An evidence-based AI chatbot designed to support emotional regulation using DBT and chain analysis.

An evidence-based AI chatbot designed to support emotional regulation using DBT and chain analysis.

Led by Dr. Liying Wang at AIMHealth Innovation Lab, GLOW was created to support the emotional regulation of patients using Dialectical Behavior Therapy (DBT) and chain analysis. The system bridges clinical insight and user-centered design to create a trustworthy AI mental health tool that complements therapy practices.

Role

UX Researcher

UX Designer

Tools

Miro, Figma & Figjam

PROBLEM

AI-powered mental health chatbots often lack clinical validation and struggle to deliver consistent, empathetic, and digestible support.

With the rise of generative AI and normalization of regular use of AI chatbots such as ChatGPT, misuse has been an increasingly large problem.

Work

Play

Resume

About

Guided Onboarding

The interactive onboarding process helps users build literacy in DBT and therapeutic language gradually, easing them into the reflective process without confusion or intimidation.

Split-Screen Interface

The split-screen layout in both the web and mobile versions enable users to view their DBT-based chat alongside the automatically generated chain analysis, providing a clear, real-time visualization of their behavioral patterns.

Editable Chain Links

Users can directly edit individual links within their chain (i.e. thoughts, emotions, behaviors) to personalize their reflections and refine accuracy over time.

Guided Onboarding

The interactive onboarding process helps users build literacy in DBT and therapeutic language gradually, easing them into the reflective process without confusion or intimidation.

Split-Screen Interface

The split-screen layout in both the web and mobile versions enable users to view their DBT-based chat alongside the automatically generated chain analysis, providing a clear, real-time visualization of their behavioral patterns.

Editable Chain Links

Users can directly edit individual links within their chain (i.e. thoughts, emotions, behaviors) to personalize their reflections and refine accuracy over time.

Guided Onboarding

The interactive onboarding process helps users build literacy in DBT and therapeutic language gradually, easing them into the reflective process without confusion or intimidation.

Split-Screen Interface

The split-screen layout in both the web and mobile versions enable users to view their DBT-based chat alongside the automatically generated chain analysis, providing a clear, real-time visualization of their behavioral patterns.

Editable Chain Links

Users can directly edit individual links within their chain (i.e. thoughts, emotions, behaviors) to personalize their reflections and refine accuracy over time.

PROCESS

PHASE 01

1

Idea Assignment

2

User Research

3

Design System Ideation

PHASE 02

4.0

Desktop Design

4.1

Usability Testing

4.2

Improvement

PHASE 03

5.0

Mobile Integration

5.1

Usability Testing

5.2

Improvement

Indx.

Boreas

Boreas

Boreas

Zephryos

Zephryos

Zephryos

Notos

Notos

Notos

Euros

Euros

Euros

Remix

IDEA ASSIGNMENT

loREM

USER RESEARCH

loREM

DESIGN SYSTEM

loREM

DESKTOP DESIGN

loREM

USABILITY TESTING

loREM

IMPROVEMENT & HANDOFF

loREM

IDEA ASSIGNMENT

loREM

USER RESEARCH

loREM

DESIGN SYSTEM

loREM