Making mental health more accessible with Winston AI

Winston is an AI therapy app for iOS designed to make mental health more accessible through personalized conversations, progress tracking, and goal-oriented coaching. I conduced competitor research, worked on wireframes and concepts, created designs and assisted with their implementation.

Goals

• Reduce the barrier to starting mental health conversations by designing a simple, low-friction onboarding experience.
• Build trust through a calm, approachable interface that encourages users to return consistently.
• Help users make measurable progress with goal setting, conversation history, and mood tracking.
• Create an experience that adapts to individual needs through personalization and AI-powered coaching.
• Validate key assumptions with beta users and iterate based on qualitative feedback.

Competitor research

I analyzed existing AI companion and mental health apps to understand common interaction patterns, onboarding flows, and approaches to personalization. I also looked at other super successful apps to learn more from their onboarding and flows. This helped identify opportunities to reduce friction, build trust early in the experience, and differentiate Winston by focusing on a conversational, goal-oriented experience rather than a traditional therapy or journaling app.

Wireframes

I created low-fidelity wireframes to explore the app's core user flows and key interactions before moving into high-fidelity designs. This allowed me to quickly iterate on the onboarding experience, conversation flow, and navigation while collaborating closely with the developer on technical feasibility.

Designs

Next up, I translated the wireframes into high-fidelity designs with a focus on clarity, accessibility, and a calm visual aesthetic. Every screen was designed to reinforce trust, minimize cognitive load, and create a consistent experience across onboarding, conversations, progress tracking, and voice interactions.

Validation

In order to test the usability and ease of use of the app, I worked with the developer to create a beta version via TestFlight, which we then shared with beta testers (users of other Solid Apps products). This helped me identify several issues I didn't previously consider:

• Some people were reluctant to give their name during the onboarding, so I added a skip button.
• Some users had specific instructions that they wanted to add, so I added a field for custom instructions.
• Interactions related to voice mode were a little awkward in the beginning, so the developer and I worked to refine this experience.

Launching the app

Once testing was done, I continued to collaborate with the developer to refine the first live version of the app, which you can now find on the App Store.

I also assisted with App Store screenshots and other assets needed to launch the app.

View on App Store

Next case study

Webfaction