FaceCart

FaceCart

\documentclass[11pt]{article} \usepackage[utf8]{inputenc} \usepackage[margin=1in]{geometry} \usepackage{amsmath} \usepackage{hyperref} \usepackage{enumitem} \usepackage{titlesec}

% Section formatting \titleformat{\section}{\large\bfseries}{}{0em}{}[\titlerule]

\title{\textbf{Facecart: AI-Powered Skincare Purchasing Assistant}} \author{Project Documentation} \date{}

\begin{document}

\maketitle

\section{Inspiration} Skincare is expensive, confusing, and risky. Many consumers depend on influencer advice or trial-and-error, which often leads to purchasing products that are mismatched, irritating, or ineffective. We built the \textbf{Facecart Web-App} to resolve these challenges by providing data-driven answers to skin health questions.

\section{What We Learned} Through development, we identified several key user insights: \begin{itemize}[noitemsep] \item Users trust systems more when they ask clarifying questions instead of making assumptions. \item Preventing a bad purchase is often more valuable than recommending a better'' one. \item Compatibility scores are more intuitive for users than binarybad/good'' labels. \item Clear explanations build more confidence than technical ingredient breakdowns. \item UX clarity is just as important as AI accuracy in purchasing decisions. \end{itemize}

\section{What We Built} Facecart is an AI-powered purchasing assistant that analyzes a user’s skin context through text input or photo analysis. Following an initial submission, the system prompts the user with clarifying questions to analyze possible causes of skin concerns. The app provides: \begin{itemize} \item \textbf{Structured Skin Profile:} A context-aware profile (not a medical diagnosis). \item \textbf{Compatibility Scores:} Treatment approaches scored from 0--100. \item \textbf{Routine Validation:} Ensures only one product per routine step to avoid over-complication. \item \textbf{Conflict Detection:} Identifies ingredient interactions (e.g., retinoids vs. strong acids). \item \textbf{Optimization:} Balances routines for both budget and effectiveness. \item \textbf{Refill Predictions:} Estimates longevity based on usage patterns. \end{itemize}

\section{How We Built It} Facecart is a layered system combining AI interpretation with deterministic validation logic:

\subsection*{AI Intake Layer} \begin{itemize}[noitemsep] \item \textbf{Natural Language Processing:} Interprets text-based skin concerns. \item \textbf{Computer Vision:} Analyzes uploaded photos for visible indicators like redness, oiliness, dryness, and inflammation. \end{itemize}

\subsection*{Skin Context & Compatibility} Raw inputs are converted into structured attributes such as barrier sensitivity and inflammation risk. The system then checks ingredient interactions, flagging conflicts with plain-English explanations.

\subsection*{Budget & Refill Logic} The system optimizes costs across the entire routine and predicts refills using the following logic: [ \text{Days of Use} = \frac{\text{Product Volume}}{\text{Estimated Daily Usage}} ]

\section{Challenges We Faced} \begin{itemize} \item \textbf{Simplification:} Translating complex skincare science into non-medical explanations. \item \textbf{Guardrails:} Balancing user freedom with safety protocols to prevent harmful combinations. \item \textbf{Tone:} Designing AI outputs that are helpful without overclaiming medical authority. \item \textbf{Accessibility:} Creating a UX that is equally functional for skincare novices and enthusiasts. \end{itemize}

\end{document}