Inspiration
Some of the most meaningful beauty and wellness advice doesn't come from an app. It comes from home: a grandmother's remedy, a mother's advice about food, or a family ritual passed down through generations.
CHARME was inspired by that idea and by a simple question: What if modern AI could help us understand when traditional wisdom is relevant to our skin today?
We also wanted to challenge the way beauty technology talks about skin. A person with vitiligo, acne, scars, pigmentation differences, birthmarks, or simply visible natural skin variation shouldn't automatically be told that something is "wrong" with them.
Our philosophy became:
CHARME doesn't judge your skin. It helps you understand it—and decide what, if anything, you want to explore.
What it does
CHARME is an AI-powered skin wellness companion that combines YouCam Skin AI, OpenAI-powered reasoning, Indian food wisdom, Ayurveda-inspired wellness frameworks, and family home remedies.
The experience follows:
LOOK → UNDERSTAND → NOURISH → RITUAL → RECHECK
Users start with a selfie, and YouCam analyzes visible skin characteristics such as hydration, redness, radiance, texture, spots, and pores.
CHARME then adds context through two lenses:
Modern Lens — what the YouCam analysis observes.
Traditional Lens — optional insights inspired by Ayurveda and Indian wellness traditions.
Users can also tell CHARME what they have in their kitchen or share a family remedy. OpenAI connects these inputs into personalized, practical guidance.
Instead of automatically telling someone what to "fix," CHARME first asks what they actually want help with: understanding their skin, exploring a cosmetic concern, building healthier habits, exploring traditional rituals, or simply learning.
How we built it
CHARME is built as a layered AI experience where each technology has a specific role.
YouCam Skin AI — Perception
YouCam provides the visual skin intelligence layer. We use its Skin Analysis API to turn a selfie into structured skin observations rather than asking an LLM to guess what it sees.
OpenAI — Reasoning
OpenAI takes the structured YouCam results together with the user's goals, food preferences, family wisdom, and selected traditional wellness context to generate personalized CHARME Insights.
Every insight actually generated by OpenAI is clearly labeled "Powered by OpenAI."
CHARME Intelligence Layer — Orchestration
Our application connects these inputs into a single workflow:
Skin observation → user intent → traditional context → food → ritual → personalized guidance → recheck
Frontend
- Next.js
- React
- TypeScript
- Tailwind CSS
- Responsive mobile-first UI
The application keeps YouCam as the visual analysis layer and OpenAI as the reasoning layer, with server-side API calls and structured outputs.
Challenges we ran into
The hardest challenge wasn't getting an API to return a result. It was deciding what an AI should do with that result.
A skin analysis can produce many measurements, but users don't necessarily want a list of numbers. They want context and a useful next step.
We also had to carefully balance traditional wellness knowledge with responsible AI.
We didn't want CHARME to blindly validate home remedies or turn Ayurveda into medical advice. At the same time, we didn't want to dismiss the cultural knowledge that inspired the product.
That led us to deliberately separate:
What YouCam observes → What traditional practices suggest → What CHARME recommends.
Another challenge was building an inclusive experience. We designed the AI prompts and UX so that visible skin characteristics aren't automatically framed as imperfections or problems that need correction.
Accomplishments that we're proud of
We're proud that CHARME became more than a skin-analysis wrapper.
We created a complete agentic experience that connects:
Computer vision + AI reasoning + nutrition + traditional wisdom + personalization.
We're especially proud of the Two Lenses concept, which allows modern skin analysis and traditional wellness frameworks to coexist without presenting them as the same type of evidence.
We're also proud of the "Just Learn" experience. A user can choose not to change anything about their appearance and simply understand their skin.
Finally, CHARME doesn't automatically send users shopping. Sometimes its recommendation can be:
You don't need to change anything right now.
That is an important part of the product philosophy.
What we learned
We learned that the most valuable part of an AI product isn't necessarily the model output—it's the translation from data into context and action.
YouCam can tell us what is visible.
OpenAI can reason about that information.
But the product becomes meaningful when we connect it to the user's actual life:
What do you care about?
What's already in your kitchen?
What did your family teach you?
What do you actually want to change?
We also learned that personalization shouldn't mean constantly generating more recommendations. Sometimes the most responsible AI recommendation is to do less.
And perhaps most importantly:
AI can personalize beauty without standardizing it.
What's next for CHARME
The hackathon version is the beginning of CHARME's larger vision.
- Human-in-the-loop dermatology
We'd build a stronger pathway for users who want professional evaluation.
CHARME could generate a structured Dermatology Brief containing:
symptom/concern timeline user-provided observations longitudinal photos relevant skin-analysis snapshots current products and routines questions for the clinician
A dermatologist would remain responsible for diagnosis and treatment decisions.
- Longitudinal skin journeys
Users could see how their skin changes over weeks and months while keeping environmental and lifestyle context alongside their snapshots.
- Deeper Indian wellness intelligence
We'd expand:
seasonal Ayurvedic-inspired rituals regional Indian food traditions multilingual Indian-language experiences a personal "family wisdom" library richer kitchen-to-meal recommendations
- More YouCam intelligence
Additional YouCam capabilities could enable richer cosmetic visualizations and personalized experiences while maintaining clear boundaries between simulation and medical assessment.
- Global cultural wellness
Longer term, we envision CHARME becoming a global platform for culturally aware skin wellness—where people can preserve the wisdom they inherited while making it relevant to who they are today.
The ultimate vision is:
AI for the first step. Human expertise for the decisions that matter.
CHARME doesn't want to replace the people we trust.
It wants to help us know when to listen to ourselves, when to learn from tradition, and when to ask a professional.
Built With
- agentic
- ai
- api
- artificial
- ayurveda
- beauty
- computer
- css
- generative
- health
- intelligence
- next.js
- nutrition
- openai
- react
- tailwind
- typescript
- vision
- wellness
- youcam
Log in or sign up for Devpost to join the conversation.