opportunity
Journaling has been proven as a powerful tool for building mental well-being. There is an opportunity to make journaling more accessible and meaningful, inviting more people to write, reflect, and cultivate self-awareness and compassion.
Background & goals
Ineyon AI’s co-founders Tobi and Keiko have a goal to build a product that helps users cultivate mindfulness and compassion. Inspired by their vision, I joined their project as the UX designer.
After in-depth research in the mental wellbeing space, Tobi and Keiko decided to create a mobile app that supports users to build mindfulness and compassion through emotion tracking, breathing and journaling. Through accessible compassionate self-reflective tools, Inyeon AI aims to cultivate the self-awareness, emotional intelligence, and meaningful connections that nurture human potential and collective well-being.
Research
I suggested that we conduct user research to pressure test our product ideas and understand what users want from a journaling experience. I decided to tackle the design for the journaling feature. I first conducted a competitor analysis of the popular journal apps. There are a variety of journal apps on the market catered to different types of journal writers.
To gain deeper insights on user behaviors and motivation for journaling, I conducted interviews (see interview questions) with 6 individuals. The interview is composed of questions on why users journal, what they typically journal about, how they prefer to journal, and what other well-being practices they perform (e.g., emotion tracking, breathing exercises).
Through user interviews, I found that journaling was widely viewed as a meaningful practice for well-being—far more than mood tracking or breathing exercises. I shared these insights with Tobi and Keiko and recommended we prioritize journaling as our core feature.
Around the same time, they consulted with an agency and shared my findings. The agency echoed my recommendation to focus on journaling first. Collectively, we decided to pivot Inyeon into a journaling app and move the other ideas to the future roadmap.
This was also June 2023, when generative AI had just begun reshaping human-AI interactions. Inspired by its potential, we decided to explore how generative AI could enhance journaling for our users. We chose Anthropic's Claude, the AI chatbot designed with safety and ethics as a priority. To inform our direction, I reviewed AI-enabled journaling tools already on the market. We also brought on a researcher, Melody, who conducted interviews with 30+ users to explore their perceptions of journaling and AI.
From this collective research, we surfaced key insights that shaped the design of Inyeon:
Getting thoughts and emotions out of the heads is the key reason people journal.
The act of writing down thoughts and feelings effectively help people process information, put things in perspectives, stop rumination, and feel relieved. All interviewees shared that they feel accomplished or relieved after completing a journal.
People run into blockers when they write down their thoughts.
Sometimes people find it difficult to dig deeper for more complex thoughts and feelings and provide enough details in the entries.
People tend to revisit their journals for various reasons.
However, for people who use physical journals or a basic digital tool (e.g., Apple Notes app), they find it challenging to look for a specific entry since there is no way to search.
People highly value maintaining a sense of autonomy when using an AI product.
They want to understand how their information will be used and protected.
Less is more.
Some people find it overwhelming when AI spits out too much information.
Design Vision & Early Ideation
A tool that helps users fully express their thoughts and deepens reflection
A safe, trustworthy place where human are in control
A stress-free, clutter-free journaling experience that fosters focus
I created low-fi wireframes to mock up the key features in Figma, focusing on the key user actions of journaling, reviewing past entries, and viewing insights. I facilitated design reviews with the team to collect feedback before moving onto high-fi wireframes.
Iteration & Testing
After multiple rounds of iterations, I developed a high-fidelity prototype in Figma and conducted virtual usability testing with three participants. Participants completed tasks by clicking through the prototype and sharing thoughts aloud. They rated the intuitiveness and helpfulness of key features on a 5-point scale. Overall, results were strong: 100% task completion, consistently positive ratings, and all participants said they would use Inyeon once launched.
Prompts are valued but may be underutilized due to discovery challenges.
All three users were confused by the lightbulb icon which was used to represent journal prompts. While prompts were seen as helpful, users had difficulty locating them.
Enhancement: Redesign the bottom navigation and relocate prompt access to the home scree to streamline the user journey. Add descriptive labels to icons, clarifying actions users can take.
Users need help with adding summary tags
Users found the summary tags helpful but would like suggestions on adding additional tags for their emotions or topics.
Enhancement: Include suggestions on tags based on journal content. Users can easily refine the summary by selecting from additional suggestions or adding their own.
Presenting data matters—but how it’s presented matters just as much. Users want to feel supported and encouraged, not just informed.
The data visualization of emotional trends was unanimously well-received. However, some concerns were raised about how negative patterns are communicated—for example, one user shared, “It’s hard to see that I mostly felt insecure over the past six months.”
Enhancement: Adjust the tone of Inyeon and how insights are presented. Introduce compassionate language, especially when presenting insights that mention negative emotions and heavy topics. Remind users the insights are solely based on user content, and do not represent the comprehensive view of users' life.
Users want more than summaries — they want AI insights to guide their reflection.
The Insight generation feature was highly appreciated for helping users recognize and reflect on their thoughts and emotions. However, they users wanted the next level insights beyond summaries: tips of recognizing thought patterns, personalized suggestions, and next steps.
Enhancement: Incorporate goal collection in the onboarding experience and provide personalized feedback related to user goals. Add a “More Insights” feature to provide more interactions between users and AI, allowing users to get tailored insights after each journal. .
I presented this feedback and the initial enhancement to the team. The team agreed with the addition of the goal collection feature. However, to my surprise, I didn’t receive immediate head nods on the “More Insights” feature.
challenges
How might we provide valuable and personalized feedback to users within each journal experience without creating an AI chat experience?
The team raised concerns around the feature of the interactions between users and AI, worrying that this steps into the AI chat territory or AI therapy, which we want to avoid. There are also potential legal and liability issues if AI “prescribes” solutions to users.
I went back to the drawing board and facilitated a Crazy Eights exercise with the team to ideate.
Among the 32 ideas that we collected among the four of us, the key themes can be summarized into 3 categories:
Encourage users to continue journaling and reflection through additional prompts and inspirations
Use visualization to trigger new perspectives and deepen reflection
Provide description and categorization to help users organize their thoughts and see things in new light
I selected ideas feasible for the current design and development, and made updates to the wireframes for the next round of feedback.
How might we provide an intuitive, scalable filter experience for users to easily locate past entries?
In addition to key word search, I want users to find past entires by summary tags (e.g., topics and emotions). The current design simply lists out all the emotion tags saved by users alphabetically, which can quickly become overwhelming.
Showing the tags under categorization is one of the ways to avoid presenting too much information all at once. However, categorizing emotions proves to be much harder.
I looked into academic research on emotion description and categorization as well as consulted other designers for inspirations. I decided to refine the filter using the emotion wheel, a well-known framework proposed by psychologist, Robert Plutchik. This design will require user testing to validate if it is an effective approach.
Wireframes
design system
Outcome & Impact
We shared the final prototype with 30 people who currently practice journaling. All expressed interest in downloading and using this app.! The team felt confident with the current design and decided to move into development.
Constraints
There is limited time to dedicate to this project. Since this is not a full-time job for any of us, I accepted that we will take the time we need to develop a Minimum Lovable Product (MLP) without burning ourselves out. Otherwise, we are not setting a good example as a startup focusing on well-being! I facilitated intentional and efficient design discussions to gather feedback from the team.
With the lean team of 5, it will take multiple releases to achieve our vision. I had in-depth feature discussions with the team and aligned on the critical features to include in the MLP. For the first launch, our scope is to build a journaling app that offers the key features that support users’ journal practice and a few preliminary AI features (e.g., journal summary, entry insights) to power up this experience.
Reflection
Balance, is the word that comes up in my mind throughout the design process.
What Users Want vs. What Users Need: Striking a balance between what users want and what users need is critical. UX professionals need to respond to users' request but also champion our product's vision and value. It is not about building every feature that users want, but thinking critically about solutions that deliver long-term, meaningful value.
Simplicity vs. Information Density: Inyeon captures a large amount of data from user entries and will use the information to provide helpful insights and inspiration to users. Features like filtering past entries highlight the challenge of creating a scalable design that maintains a clutter-free interface while providing users access to rich, personalized content.
Empowerment vs. Dependence: Inyeon's mission is to facilitate genuine, deep relationships with ourselves and with others. While we aim to create a product users enjoy and trust, we must be mindful not to encourage dependence. AI should support, not replace, meaningful human interactions. Every design decisions must weigh short-term delight against long-term well-being to make sure the Ineyon experience is safe, ethical, and human-centered.