
Evolve is a mental health startup aimed at providing accessible tools for emotional well-being.
With over 1 million users worldwide, Evolve has been recognized for its impact in the mental health space. Featured by Apple across 82 countries and awarded the prestigious Apple Editors’ Choice award, it continues to shape how people integrate mental wellness into their daily lives.
Understanding the problem
“Ever had a day where everything feels off, and you just need someone to listen? At Evolve, we noticed many users logging 'sad' moods but not finding the support they needed afterward.”
Our mood tracker showed that 62.5% (30 days data) of users reported feeling sad. With more than 500 users logging there mood more than one time.
However, the app's response was a generic confetti animation, which felt more like a celebration than support.

Problem discovery, UX research, product thinking, user flow design, interaction design, cross-function collab
- Tanmay Goel- iOS enginner
- Surbhi Jain - AI engineer
- Shiv Kumar - Backend devloper
- Deeptanshu - Backend intern
Jan 2025 - Feb 2025
The Current Flow

The Flow
- 😶Clicks on the mood tracker
- ☹️Selects a mood
- 🥺Selects an emotion they are feeling
- ✍️Reflect about there mood
- 🎉Save takes you to the home screen with confetti and task completion state
Right now, our flow feels like we are celebrating their sad emotion.
Instead, in reality, if a friend is not feeling well, we would try different ways to make them feel heard. And after the conversation, we would ensure they feel better than before.

Why do we journal?
- To reflect.
- To come to a conclusion.
- To feel light.
- To document our life moments

Just to understand this more, I also interacted with therapist.
I was very curious to understand how they help users, to which she answered
“We don’t give them suggestions or answers—we help them find their own answers by reflecting.”
This insight resonated with me, so I pitched this problem statement and took the lead on solving it.
For this sprint, as version 1, I wanted users to feel heard and help them reflect on why they are feeling this way.

Ideation
To improve the sad emotion journaling experience, I started by looking at features we already had — like affirmations and chat with therapist.
“How can we help them feel heard and feel slight better —without overwhelming or overstepping?”
Suggestions with Affirmations
What if we give an affirmation just after they have completed there journal entry.
The hypothetisis was that people who are feeling sad will get a hope after they read a postive comforting message that might make them feel positive.
The issue:
- figuring out which affirmation to show in its own is another task
- recommending some random affirmation might not resonate with them
- I am not sure if it will replace the need to reflect more
- choosing the right one? Not that easy.

Reflect more with AI
How about we train AI to help users reflect more, by asking the right questions.
The team loved this idea too — since they were already exploring how AI could support mental health features
The Pros:
- Scalable and always available
- Doesn’t try to “fix”—just listens and gently reflects
- Strong potential for long-term growth and monetization

We decided to move forward with “Reflect with AI” this approach for V1, and see how users respond before expanding it further.
Final designs and the flow

The “Journal History” flow
We added the AI feature to all the three journals of evolve - Mood tracker, Today’s journaling and Gratitude journal

We launched this version on Jan (the before) and on Feb (the after), there was 5% increase on journal click because of just adding “reflect with evolve AI” text.
The edge cases
Suicidal or self-harm mentions
Redirect to a dedicated crisis helpline screen with immediate resources.
The Impact
User Adoption:
Since launching on February 10th, our AI chat feature has seen promising engagement, with 20.6% of daily journaling users (just one week after launching) utilizing it. This indicates users find value in exploring their thoughts with AI post-journaling.
Journal click open:
There was 5% increase on journal click because of just adding “reflect with evolve AI” text.
Usage Patterns:
We've identified power users engaging significantly, with one user reaching 18 interactions in a day. On average, users engage with the AI 7 times, with each chat session averaging 1.5 interactions.
Growth Opportunities:
While existing journal power users haven't fully adopted AI chat yet, this presents an opportunity to refine our integration and user education strategies. These early metrics provide valuable insights for enhancing user experience and driving deeper engagement across our platform.