Our project was inspired by the common struggle people face when it comes to skincare—choosing the right products, understanding their skin type, and dealing with issues like acne, pigmentation, or oiliness without proper guidance. Dermatologists aren’t always accessible or affordable, and online advice is often confusing or contradictory. We wanted to build an AI-powered solution that makes skincare more personalized, affordable, and easy to understand.

To build this, we combined computer vision and machine learning to analyze skin concerns from selfies and generate tailored recommendations based on skin type, lifestyle, and even local climate factors such as humidity, UV index, and pollution. We also added a simple chatbot-style interface that allows users to ask skincare-related questions and receive AI-driven responses in real time. Our tech stack included Python with TensorFlow for the analysis model, FastAPI for backend support, and a lightweight frontend for a smooth user experience.

Throughout this project, we learned how AI can be applied to real-world personal care problems, how important it is to balance accuracy with user trust, and how to scope an MVP that delivers value under hackathon time limits. One of our biggest challenges was training the skin analysis model with limited data and ensuring it worked across different lighting conditions, camera qualities, and skin tones. Another was designing a user experience that felt helpful while staying safe and general enough, since we are not certified dermatologists. Despite these challenges, building this project taught us the importance of merging technical innovation with everyday usability, and it showed us how AI can truly make an impact in people’s daily lives.

Built With

  • along-with-javascript-(react)-to-create-a-clean-and-user-friendly-web-interface.-on-the-machine-learning-side
  • and-pollution.-the-application-was-hosted-on-google-cloud-for-scalability
  • and-recommendations.-to-enhance-personalization
  • fastapi
  • humidity
  • javascript
  • opencv
  • python
  • scikit-learn
  • skin-progress
  • tensorflow
  • we-implemented-computer-vision-techniques-using-tensorflow/pytorch-for-skin-analysis-and-opencv-for-preprocessing-images-to-handle-lighting-and-quality-variations.-the-backend-was-developed-using-fastapi
  • we-integrated-external-apis-such-as-openweathermap-to-factor-in-real-time-climate-conditions-like-uv-index
  • we-used-python-as-the-core-language-for-building-the-ai-models-and-backend-logic
  • web-app
  • which-allowed-us-to-serve-the-model-efficiently
  • while-mongodb-was-used-as-our-primary-database-to-store-user-profiles
Share this project:

Updates

Submission history