CareFi

AI powered skincare platform that analyzes your skin from photos, and detects key traits. Along with personalized, budget-friendly routines with product recommendations and savings.

Comment 2

Inspiration

We were inspired by a friend’s dermatology club that asked us to create a website for skincare analysis. This sparked our idea to build an AI-powered platform that makes personalized skincare accessible and affordable for everyone.

What it does

CareFi uses advanced AI to analyze a user’s skin from photos and detect common conditions such as acne, dryness, oiliness, and sensitivity. It then recommends a customized skincare routine tailored to each user’s unique needs and budget, providing insights often associated with dermatologist-quality results.

How we built it

We utilized a modern, full-stack approach.

  • Frontend/Backend: Next.js for high performance, API and App routing, SSR, and TypeScript + TailwindCSS

  • Database: PostgreSQL (Supabase) for highly consistent & horizontal data storage, user authentication, RLS policies for security, and real-time subscriptions / webhooks

  • AI + Analysis: We used OpenAI Vision (specifically gpt-4o-mini) to analyzing uploaded images. We have another agent for also the recommendation process too. Out of a list of 100 products, we send the agent 40 unique & well matched products (tier-based classification based on ingredients list) to the agent and have it write the list of recommended products and measure through a confidence scale.

Architecture: RESTful API routes in /app/api/ Endpoints include:

  • /api/signup - User registration
  • /api/signin - User authentication
  • /api/analysis/start - Start face analysis
  • /api/analysis/latest - Get latest analysis results
  • /api/recommendations - Get personalized recommendations
  • /api/uploadImage - Upload face images
  • /api/settings/* - User settings management

Hosting: Vercel - Serverless deployment platform. Automatic builds on push. Edge network CDN. Environment variable management

Key Features Built with This Stack

  • Product Image Analysis - Upload photos → AI extracts ingredients
  • Personalized Dashboard - View analysis, recommendations, KPIs
  • Budget Optimizer - Track spending on skincare products
  • Routine Planner - Morning/evening skincare routines
  • Onboarding Flow - Collect user skin profile (type, concerns, allergies

Challenges we ran into

We ran into the start of being a new team and this being our first hackathon, we were unfamiliar with what goes into a hackathon. Integrating our AI.

Accomplishments that we're proud of

We are proud of how much we learned and were able to accomplish in such a short amount of time. We are also proud of our real-time analysis of skin along with actively fetching and updating the user's data.

What we learned

As a team participating in our first hackathon, our journey was marked by significant but rewarding challenges. We immediately encountered a steep learning curve in managing our collaborative workflow and project coordination under intense time pressure. Technically, the most demanding task was successfully integrating the fine-tuned AI model with our Next.js web framework while simultaneously ensuring real-time performance for the skin analysis feature.

What's next for CareFi

For the future of CareFi, our next immediate step is to partner with the campus dermatology club to deploy the application for a pilot program, allowing us to gather crucial real user feedback. Moving forward, we plan to continually expand our AI model's capabilities, enhance the quality of product recommendations, and work toward continuously improving the overall accuracy of our skincare insights.

Built With

Share this project:

Updates