Alina

Your AI Skin Companion

Comment

Inspiration

Skincare advice today is often generic. People buy products based on trends, social media, or marketing without understanding what their skin actually needs. Even when they use skin analysis apps, they usually receive a one-time report with little context and no long-term guidance.

I wanted to build something that feels like a personal skincare companion rather than a simple scanner. Instead of just analyzing a face, Alina remembers previous scans, understands how a user's skin changes over time, and provides personalized recommendations that evolve with them.

My goal was to combine professional skin analysis with an intelligent AI agent that reasons about the user's history instead of generating generic responses.

What it does

Alina is an AI-powered skincare assistant that analyzes a user's skin, tracks progress over time, and generates personalized skincare guidance.

Users upload or capture a facial image, which is analyzed using Perfect Corp.'s YouCam Skin Analysis API. The scan detects dozens of skin attributes, including skin age, acne, wrinkles, pores, redness, oiliness, moisture, firmness, texture, radiance, dark circles, and more.

After the scan is completed, a Google Gemma 4 AI Agent takes over. Rather than simply summarizing the results, the AI:

  • Retrieves previous scans and historical observations.
  • Compares the current scan against past results.
  • Generates easy-to-understand summaries.
  • Creates detailed observations.
  • Suggests personalized skincare routines.
  • Recommends products tailored to the user's skin.
  • Suggests healthy skincare habits.
  • Answers skincare questions through a conversational AI assistant.

Every recommendation is based on the user's skin data and previous history, allowing guidance to improve over time instead of treating every scan as an isolated event.

How we built it

I built Alina as a modern full-stack web application.

The frontend was developed with React and TypeScript, while the backend uses Fastify, TypeScript, PostgreSQL, and Prisma ORM.

The AI workflow combines two major technologies:

  • Perfect Corp. YouCam Skin Analysis API performs professional facial skin analysis and returns detailed skin metrics.
  • Google Gemma 4 powers an AI agent capable of reasoning through scan results using function calling.

Instead of relying only on the AI's context window, the agent can retrieve previous scan history, recommendations, routines, observations, summaries, and user preferences through function calls before generating a response. This allows Alina to make informed decisions based on long-term skin progression rather than a single scan.

Challenges we ran into

One of the biggest challenges was designing an AI agent that behaves more like an assistant than a chatbot.

Instead of allowing the model to immediately answer questions, I had to build an orchestration layer that lets the AI determine which information it needs, execute multiple function calls, process the returned data, and only respond after gathering enough context.

Another challenge was transforming raw skin analysis values into explanations that are meaningful for everyday users. Numerical scores alone don't help people understand their skin, so i focused on producing personalized, plain-language insights.

Integrating advanced skin analysis with a reasoning AI while keeping responses fast and contextual also required careful backend architecture.

Accomplishments that we're proud of

I am proud of creating an AI agent that goes beyond answering questions.

Alina can retrieve its own knowledge before responding, compare historical skin data, generate structured observations, recommend personalized skincare routines and products, and provide guidance that improves over time.

I am also proud of integrating professional skin analysis with an intelligent reasoning workflow into a seamless experience that feels like talking to a personal skincare consultant.

What we learned

This project taught me that building AI products involves much more than prompting a language model.

I learned how to design an AI agent capable of reasoning through multiple data sources, orchestrate function calling, structure long-term user memory, and transform complex analysis into clear, trustworthy explanations.

I also gained valuable experience integrating enterprise AI APIs into a production-ready application.

What's next for Alina

This hackathon version is only the beginning.

Next, i plan to add real-time facial scanning, mobile applications for Android and iOS, voice conversations with the AI assistant, smarter product recommendations, ingredient compatibility checking, routine adherence tracking, and deeper trend analysis across months of skin history.

I also want Alina to become a complete personal skincare companion that proactively helps users build healthier skincare habits instead of only reacting after a scan.

Built With

Share this project:

Updates

Submission history