SkinIQ

Logo

Inspiration

Waking up one day and realizing how challenging it is to find a personalized, effective skincare routine sparked the idea for SkinIQ. I was struggling to clear my ance and always wanted a more detailed knowledge resource telling me what to do with my skin, hence SkinIQ. Our goal was to create an intelligent assistant that simplifies skincare decisions by leveraging AI and web-based recommendations.

What it does

Simple as it sounds, IQ for your skin, so you know everything about your skin with a click of a picture!

There are four main elements to our app:

Skin Type Detection – Analyzes user inputs to determine if their skin is oily, combination, or dry.

Daily Routine Generator – Suggests a tailored morning and night skincare routine.

Skincare Chatbot – Answers user queries related to skincare concerns, product compatibility, and best practices.

Routine Generator with Product Recommendations – Searches the web to recommend suitable skincare products available for purchase.

How we built it

We built it with a variety of elements some of them being: Python, Scikit-learn, Supabase, PyTorch, Keras, TensorFlow, Vercel, TypeScript, React, FastAPI, GitHub, OpenCV, Google Auth, Kaggle, and Canva

Challenges we ran into

Data Accuracy: Ensuring reliable product recommendations required filtering misinformation and outdated web sources

Skin Type Analysis: Developing a model that accurately determines skin type based on user responses was complex

Scalability: Optimizing API calls and handling multiple user interactions without latency issues

Personalization: Balancing general skincare guidelines with tailored recommendations for individual needs

Accomplishments that we're proud of

Successfully integrated ML Model for multiple disease detection with a connection to an LLM-powered skin type into the skincare routine and informative chat

Implemented a web-based product recommendation system that provides real-time skincare suggestions

Developed an intuitive, user-friendly website interface that simplifies skincare decision-making and learning about YOUR Skin’s IQ

What we learned

There were plenty of new things we learned in the 24 hours, from DJing (yes we took turns when we needed a break) to how to properly deploying our final product. But jokes aside, the importance of high-quality data in building reliable AI models, web scraping techniques for dynamic product recommendations, the user experience considerations for chatbot interactions in the skincare industry, and the significance of personalization in consumer applications are just some of the knowledge we gained during this 24-hour time frame!