Inspiration
We were inspired by a shared 2 AM realization at ICHack. Looking around at our peers (and ourselves), we noticed a common pattern: exams and deadlines cause stress, stress leads to hair loss, and noticing hair loss creates even more panic. Many of our friends worry about receding hairlines but are too busy or intimidated to visit a clinic. We wanted to build something that bridges vanity (looking good) and sanity (mental wellness) in a way that Gen-Z actually enjoys.
What it does
Hair Me Out estimates hair coverage from user-uploaded photos and tracks hairline changes over time. It combines AI hair analysis with a gamified wellness experience, encouraging better sleep, hydration, and stress management. Users also interact with a “Hair Fairy” — a sassy but supportive AI companion that adapts its tone based on stress levels.
How we built it
We built Hair Me Out using a full-stack approach. The mobile frontend was developed with React Native. Hair analysis was implemented using Python and OpenCV for image segmentation. The AI “Hair Fairy” persona is powered by an LLM with prompt engineering to dynamically balance sass and empathy based on user stress data. We also designed a cohesive, minimalistic 3D visual style for the avatar.
Challenges we ran into
Accurately segmenting hair and skin under different lighting conditions was technically challenging. We also had to carefully tune the AI’s personality to make sure the “roasts” were funny and motivating without being hurtful. Integrating the avatar visuals and stabilising AI behaviour within a short hackathon timeframe added further complexity.
Accomplishments that we're proud of
We built a fully working prototype in just 24 hours. The app can analyse hair coverage from images, track trends over time, and deliver context-aware feedback that switches between humour and emotional support. We are especially proud of the strong visual identity, which makes a medical topic feel fun and approachable.
What we learned
We learned that medical and wellness technology doesn’t have to be boring — emotional connection and personality significantly improve user engagement. We also gained a deeper appreciation of the challenges involved in computer vision, particularly when working with unstructured selfie data.
What's next for Hair Me Out
Next, we plan to integrate wearable data (such as Apple Health and heart rate variability) to automatically detect stress and trigger support mode. We also want to add AR features to visualise hairstyles and eventually partner with dermatologists to connect users with real professionals when serious issues are detected.
Log in or sign up for Devpost to join the conversation.