Inspiration
As someone who’s seen colleagues and loved ones struggle with unmanaged anger—either exploding outward or suppressing it inward—I knew there had to be a better way than traditional therapy alone. I wanted to create a tool that meets people in the moment of anger, not just in a weekly session, and turns that raw emotion into something actionable.
What it does
AngerTransform is a digital platform that guides users to log anger episodes in real time, access 10-second grounding tools for immediate calm, and use AI to identify recurring triggers. It also walks users through setting healthy boundaries, channeling anger into positive change, and practicing nonviolent communication (NVC) for difficult conversations—all while tracking progress over time.
How we built it
We started with user research to map core pain points (speed of access in crisis, personalized insights, actionable steps). The frontend was built in React with Figma-designed minimal UI (critical for high-stress moments), backend on Node.js/Express, and MongoDB for journal data storage. TensorFlow powered the AI pattern analysis, and AWS handled weekly/monthly report generation. We prioritized one-tap functionality for grounding tools and logging to reduce friction.
Challenges we ran into
The biggest hurdle was balancing AI complexity with simplicity—users in anger don’t want to navigate clunky analysis. We also had to ensure offline functionality (local storage for journal entries) for moments when internet access is spotty. Testing the grounding tool’s timing (10 seconds exactly) and animation smoothness across devices was another tough lift.
Accomplishments that we're proud of
We created a tool that’s both evidence-based (rooted in NVC and anger management research) and user-centric—beta testers reported 30% faster calm during anger episodes with the grounding tool, and 60% found AI trigger insights “eye-opening.” The milestone celebration feature also boosted consistent use (80% of testers logged 10+ episodes).
What we learned
Emotional tools demand extreme empathy in design—every button, animation, and word matters when a user is upset. We also learned that AI insights only work if they’re actionable (not just data dumps) and that progress tracking needs to be encouraging, not shaming.
What's next for this project
We plan to add short, audio-guided grounding exercises (for hands-free use), integrate with calendar apps to flag high-risk times (e.g., stressful work deadlines), and expand NVC practice scenarios to cover more cultural contexts. We’re also exploring a tiered subscription model to make premium features accessible to more users.
Built With
- amazon-web-services
- firebase
- tensorflow
- typescript