Meet the team!
My team consisted of 5 individuals:
-
Me - Lead Researcher + Co-designer
-
Vincent - Product Manager
-
Ryan - Researcher
-
Miguel - Lead Designer
-
Kyle - Lead Developer
Defining the problem
The demand for mental health services is at an all-time high, with approximately 56% of College students actively seeking or wanting to access these services.
​
However, as the demand increases, so do the barriers to obtaining mental health support. With guidance from UX professionals at IBM, we were challenged to design a mobile app that incorporates AI to help alleviate these barriers.
How I approached research Pt.1
My role for this project was to combine qualitative and quantitative data to discover insights on who our users are, what do they need, where do they come from, why do they need this, and how do we solve their problems?
​​
I conducted 50 student interviews and received over 100 survey responses.
​​
The findings concluded that many don't trust AI with their personal information due to uncertainty about where their data is being stored and how it is utilized.
Another insight was that many are unsure about HOW to access resources and WHAT they need for help.
The Research Pt. 2
Once I had established the who, what, where, and why, the next step was to discover how.
​​
There is an AI model called "HAILEY" that formulates empathy based responses, demonstrating that empathetic textual replies increase human trust in AI.
​
Cute mascots have proven effective in building trust due to their positive demeanor, causing individuals to feel more comfortable and secure.
Journaling has been shown to enhance mental stability by facilitating emotional expression, which relieves stress and anxiety, enhances clarity, and fosters a positive mind set.
Our Competition
As of now, there aren't many "AI-powered" mental health applications, but there are many "Non-AI-powered" apps.
​
We gained inspiration from a couple of apps: Voidpet Garden and Finch. However, we found weaknesses, such as pricey gamification, overwhelming or underwhelming amounts of features, and often quite complicated user interfaces.
​
Overall, most mental health applications felt like they had a high barrier to entry.
Meet our Persona!
Introverted Isabella is an 18-year-old college student who struggles with social anxiety and depression. She wishes she could improve her social life, feel better about herself, and do well in school, but her anxiety makes it hard to talk to people, and her depression often leaves her stuck in bed, unable to focus in class. She wants to get help but doesn't know where to start or what resources to turn to.
User Journey
Our user begins their journey feeling rather uncertain and experiencing a high level of negative emotions. ​
​​
Recognizing this, we designed our app with an empathetic and supportive approach, ensuring a low barrier to entry.
Prototyping
During the initial design process, we went through low-fidelity prototypes, user testing, and established what we wanted as our overall brand identity.
​
The mascot was designed by me! :)
Design Pt.1: Onboarding
The part you all have been waiting for: The Final Design. The onboarding process is quick and simple, with the only difficult choice for the user being which pet they want as their AI partner. Users can then input some optional data they wish to share, which will not interfere with their experience besides personalization.
Design Pt.2: Journaling
Now onto the main course... Journaling. Users can write down whatever they want in their journals on the app or in physical form, which they can then scan and upload to their digital journal, and format it however they'd like!
​​
These entries can be used either generate daily tasks that they can schedule based on keywords picked up by the AI or have a topic of discussion if the user wishes to talk to the chatbot about the journal entry they submitted.
​
All data is stored strictly by us, and users have the option to opt out of AI features if they have privacy concerns. In that case, the app instead provides generally healthy daily tasks for the user to do.
Design Pt.3: Gamification + Tasks
For engagement and to provide a more enjoyable approach to getting healthier, we decided on a gamified solution.
The AI-generated tasks will not only award users points for completing their tasks but also allow them to delve deeper into facts and reasoning behind the tasks and how they were generated or selected.
​​
With the points collected from completing these tasks, users will be able to climb tiers in a "Battle Pass" system. In this system, as users climb to higher tiers the rewards become better. Most of these rewards are cosmetic, thus optional, and don't affect the app's experience in any way. If users are willing to spend money, they can upgrade their Battle Pass in order to gain more "visually appealing" rewards.
Design Pt.4: AI Chat Bot
Referring back to earlier research on "HAILEY," the AI chatbot model that provides empathy-based responses, our AI chatbot will utilize data from user journal submissions(with their approval). This data will enable the chatbot to generate personalized conversations and advice, acting like a supportive friend available to the user 24/7.
​​
If users choose not to share journal data, the chatbot will still function with standard responses and conversation starters.
Design Pt.5: Resources
Articles, videos, and resources from credible sources such as medical journals, professionals, and research papers will be provided to the user based on their journal entries and chats.
​​
If users have AI features disabled, the resources tab will still provide useful general mental health information from reliable sources.
Key Performance Indexes
The main goal is to provide a tool to users in order to improve their quality of life. As long as user reviews are positive and activity on the app is consistently high, we can confidently say that the app is a success.
Conclusions and Next Steps
Our solution wasn't the first of the many we came up with. It took countless retries, testing, and research to find a proper solution to the issue at hand.
​​
Understanding creativity within constraints was crucial. We had to recognize that although many solutions could be envisioned, it is not possible to implement them all at once, at least not effectively.
Thank you!
Feel free to reach out to me on either LinkedIn or simply email down below if you have any questions or opportunities for me! :)
