Description:
How I designed the first agentic AI beauty assistant for girls that remembers your skin, matches your skin tone, undertone and personalizes every recommendation just for you.
The Problem Nobody Is Talking About:
You pick up your phone. You just scroll for hours finding 47 foundation shades. You select the one that appears directly on your screen. You have to wait four days for the delivery. When you try it on after opening the package, your face appears gray. or orange. or entirely incorrect.
You're not by yourself.
In India, 40% of foundation purchases are returned because the shades don't match. An estimated 35 to 40% of foundation orders in India's beauty e-commerce market are returned annually. Additionally, there is no way for millions of users in non-metropolitan areas without physical beauty stores to test products before making a purchase. Every purchase is an estimate. Money is spent on each guess. Confidence is lost with every incorrect shade.
The solutions that are currently in place are completely ineffective. Your undertone, skin type, past reactions, and the fact that your skin in December differs greatly from your skin in June are all ignored by camera-based shade matching apps, which provide you with a generic result. As soon as you exit the app, they forget about you. They give everyone the same product recommendations.
This has a word. We refer to it as the "wrong shade syndrome." And it has been unresolved for far too long.
The solutions that are currently in place are completely ineffective. Your undertone, skin type, past reactions, and the fact that your skin in December differs greatly from your skin in June are all ignored by camera-based shade matching apps, which provide you with a generic result. As soon as you exit the app, they forget about you. They give everyone the same product recommendations.
This has a word. We refer to it as the "wrong shade syndrome." And it has been unresolved for far too long.
Introducing GlowMind:
There is a perfect match for every skin tone. Each person has a distinct glow.
GlowMind isn't a chatbot. It is not a filter that matches shades. It's not just another generic beauty recommendation system.
GlowMind is a fully agentic AI system that protects you from dangerous combinations, remembers your skin, thinks before speaking, and gets smarter with each conversation you have with it.
It was based on a single idea. Although general AI tools are excellent, they are not specifically tailored for anyone because they are designed for everyone. Because GlowMind was created specifically for your skin, it is unique.
The Concern In Detail:
It's crucial to comprehend the breadth of GlowMind's problems before learning what it does.
There is such a thing as Wrong Shade Syndrome. Under different lighting, camera settings, and screen calibrations, foundation, blush, and concealer appear significantly different. Under office lighting, a warm beige that appears flawless on your phone screen may appear ashy and incorrect on your skin. Camera-based matching does not capture the true nature of your skin; it only records what the camera perceives.
The silent killer of beauty e-commerce is a lack of personalisation. Current applications rely on generic databases and one-time camera scans. The last time you used a fragrance-heavy foundation, they were unaware that you had breakouts. They are unaware that benzoyl peroxide dries out skin while niacinamide works wonders.There is such a thing as Wrong Shade Syndrome. Under different lighting, camera settings, and screen calibrations, foundation, blush, and concealer appear significantly different. Under office lighting, a warm beige that appears flawless on your phone screen may appear ashy and incorrect on your skin. Camera-based matching does not capture the true nature of your skin; it only records what the camera perceives.
The silent killer of beauty e-commerce is a lack of personalisation. Current applications rely on generic databases and one-time camera scans. They do not know that you broke out the last time you used a fragrance-heavy foundation. t's crucial to comprehend the breadth of GlowMind's problems before learning what it does.
There is such a thing as Wrong Shade Syndrome. GlowMind was designed to solve all three of these issues at once.
What Makes GlowMind Different — The Four Novelties
GlowMind does four things that no other beauty app currently on the market does.
Persistent Memory
As soon as you shut down any beauty app you've ever used, it forgets you. GlowMind will always remember you. A structured SQLite database is used to permanently store your skin tone, undertone, skin type, concerns, previous products, reactions, budget, and routine. GlowMind loads your entire history and resumes where you left off each time you come back. It keeps a dynamic, changing profile of your skin that gets richer with each interaction, rather than just remembering your conversations.
Reasoning Chain for ReAct
The majority of AI systems operate in a mysterious manner. You pose a query. There's a response. You don't know why. GlowMind is unique. The ReAct pattern—Reason, Act, Observe—is adhered to. It thinks aloud, step-by-step, before making any recommendations. It explains what it looked at, what it discovered, and why it is recommending a specific product. Each suggestion is completely traceable and explicable.
Identification of Ingredient Conflicts
This is the first of its kind in any beauty application and GlowMind's most protective feature. Vitamin C and retinol should never be taken together because they irritate one another and lessen each other's benefits. When combined, retinol and AHA seriously harm the skin's barrier. Retinol is totally deactivated by benzoyl peroxide. The majority of people are unaware of this. This is something that most beauty apps don't tell you. Instead of alerting you after the damage is done, GlowMind proactively warns you before you combine incompatible actives.
Seasonal Skin Drift
Your skin does not look the same in January as it does in July. Throughout the year, your skin type and tone are influenced by temperature, humidity, sun exposure, and lifestyle changes. GlowMind automatically asks you to update your profile if it is older than six months. Your skin's natural seasonal changes are not taken into account by any other beauty product.
Pipeline:
GlowMind uses a six-stage pipeline to process each query.
User input, memory recall, LLM reasoning, tool execution, live search, and personalised response.
Your input, which could be a text message or a single selfie taken during onboarding, is where the journey starts. GlowMind loads your entire persistent profile from the database as soon as you speak. Before it even starts to think, it already knows who you are.
During the reasoning phase, your query is considered in relation to your unique skin profile by Groq's robust language model, which is coordinated through LangChain. It chooses which tools to call and in what sequence.
Three tasks are carried out by the tool's execution: BERT sentiment analysis on actual product reviews, ingredient conflict checking against a curated database of hazardous active pairs, and shade matching against 9000 actual Sephora products.
Every recommendation comes with actual purchasing information because SerpAPI's live search retrieves current prices and availability from the internet.
Lastly, GlowMind provides a friendly, well-reasoned, detailed recommendation that stays within your budget, explains each option, and alerts you to any conflicts.
Technology Stack:
Groq LLM - llama-3.3-70b-versatile - the LLM that does all the natural language reasoning and step-by-step thinking.
LangChain - the orchestration framework that connects the LLM to all tools, memory, and data sources and implements the ReAct agent pattern.
Hugging Face CLIP - OpenAI's vision language model that recognizes skin tone automatically in a selfie by matching the image to eight text descriptions of skin tones.
Hugging Face BERT - reads product descriptions and reviews in the Sephora dataset and gives sentiment scores to prevent GlowMind from recommending a product with low reviews.
Kaggle Sephora Dataset – This contains 9000+ real beauty products with shade codes, ingredient lists, ratings, reviews, and prices. This is used as the database for RAG retrieval.
SerpAPI – This gives GlowMind the ability to search the internet for current product prices and availability for Indian websites such as Nykaa and Myntra.
SQLite – This is a lightweight database that stores every user's skin profile permanently across sessions.
Gradio – This is an interface that turns all the Python code into a beautiful shareable web app with a dark elegant theme matching the GlowMind brand.
User Experience :
Opening GlowMind for the first time is nothing like opening a beauty app.
No generic quizzes.
No popup ads.
No generic app recommendations waiting for you on your home screen.
New users have two ways to get started.
The text form option allows you to input your skin type and concerns manually.
The selfie option uses CLIP’s AI to detect your skin tone from a single selfie.
Regardless of which option you choose, your profile will be saved and remembered forever.
Returning users simply need to input your User ID.
Your entire skin history will be loaded instantly and displayed on the left panel.
Then, and only then, does the conversation start.
You type normally.
You ask questions like you would ask a knowledgeable friend.
What foundation should I get that’s under 800?
Is my current routine safe to use together?
What should I use for dark spots?
GlowMind thinks about your profile and calls its tools and answers your questions in a warm and personal step-by-step fashion that uses your name and stays within your budget.
Future Work:
The next release will also include computer vision for shade detection, allowing for a user's skin tone's exact hex color code to be matched directly with the hex codes of product shades for pixel-perfect accuracy. A mobile app will allow GlowMind to be available on every mobile phone in India, providing personal beauty intelligence for the user in every city and town. The Indian brand list will also be expanded to include Lakme, Sugar, Colorbar, Mamaearth, and all other brands of importance for the Indian beauty user.
The mission is the same. Every skin tone has a perfect match. Every person has a unique glow. GlowMind is dedicated to finding it.
About:
GlowMind is conceptualized and being created by Ayushi Shukla as a part of the BCA seminar project. It is the intersection of agentic AI, personalization, and real-world beauty tech.
Built with Groq, LangChain, Hugging Face, Sephora Data, SerpAPI, Gradio.
Match Your Magic.
Top comments (0)