Designed & Built by Jeovany GutierrezGitHub: github.com/JeoHelps/curlcue-portfolio
CurlCue
Product intelligence for curly hair
Scan. Understand.Choose better.
Built a science-backed product intelligence system for curly hair using React Native, TypeScript, and Supabase — combining barcode scanning, a 100+ ingredient database, and profile-driven scoring across curl type, porosity, thickness, and goals.
🔬 Science-backed analysis • 📱 Mobile-first • 🧪 100+ ingredients mapped
Hair Profiles
9 curl types · 3 porosity levels
Ingredient Database
100+ ingredients · 7 signal types
Routine Roles
8 categories with coverage analysis
CurlCue Scanner
Barcode-first curly hair analysis
Scan result
Curl Enhancing Smoothie
Impact
Curly-hair product decisions are still guesswork.
- •Built a science-backed scoring engine with 100+ ingredients mapped to hair-specific signals
- •Designed a 5-step profile system covering curl type, porosity, thickness, scalp type, and goals
- •Implemented role-aware scoring across 8 routine categories with confidence metrics
- •Created a full product pipeline: scan → analyze → score → save → compare → recommend
Ingredient lists are dense, recommendations are inconsistent, and most advice is not tailored to the individual. That makes finding the right product slow, expensive, and frustrating.
The solution
CurlCue replaces guesswork with science.
Scan a product, map ingredients against a research-backed database, and generate a profile-aware score tied to curl type, porosity, thickness, scalp type, and goals. Then organize it into a role-based routine with coverage analysis.
Core Systems
Core systems powering every product decision.
Science-backed ingredient analysis, profile-driven scoring, and a Routine Builder with coverage tracking — designed for real-world haircare decisions.
Barcode Scanning
Instant product lookup with Expo Camera — scan in-store or at home for real-time ingredient analysis.
Science-Backed Analysis
100+ ingredients mapped to signals like moisture, protein, drying, and buildup using peer-reviewed research on hair chemistry.
Profile-Aware Scoring
Products scored against your curl type, porosity, thickness, scalp type, and goals with confidence metrics and role-based weighting.
Routine Builder
Organize products across 8 routine roles with coverage analysis to identify gaps in your haircare routine.
Technical Decisions
Research-backed, benchmark-tested.
- •Chose React Native with Expo SDK 54 and TypeScript for type-safe cross-platform development
- •Built scoring logic backed by peer-reviewed research on protein molecular weight and curl biomechanics
- •Designed Supabase schema for product → ingredient → analysis → recommendation relationships
- •Implemented benchmark testing suite to validate scoring accuracy across hair profiles
Scan a barcode with Expo Camera
Map ingredients against 100+ classified compounds
Score with profile-aware, role-weighted logic
Save to routine, compare, or get recommendations
What I’d Do Next
Ingredient-level precision. Not generic advice.
- •Expand product database through user-submitted product requests and admin curation
- •Refine scoring weights with additional hair science research and user feedback
- •Add routine history tracking for trend-based recommendations over time
- •Deploy to TestFlight and run benchmark validation across real user profiles
Product experience
Designed like a real mobile product.
A full flow from scan → analyze → routine → compare. Built for real usage, not just concept.
Home Dashboard
Personalized recommendations, routine insights, and quick actions — all driven by your hair profile.
Routine Builder
Organize products across 8 role categories with coverage analysis to surface gaps in your routine.
Product Details
Science-backed ingredient breakdown with profile-aware scoring, signal mapping, and role inference.
Scanner
Expo Camera barcode scanning with real-time product lookup and instant ingredient analysis.
Profile
5-step onboarding covering curl type, porosity, thickness, scalp type, and goals to drive scoring.
Compare
Side-by-side product comparison ranked by compatibility with your specific hair profile.
Tech stack
Built with a type-safe, research-driven stack.
TypeScript for type safety, Expo SDK 54 for cross-platform development, Supabase for scalable data, and a custom scoring engine backed by peer-reviewed hair science research.
About the builder
Built independently as a production-style product — not a tutorial or class assignment.
Jeovany Gutierrez
UC Merced Computer Science & Engineering student focused on building science-backed mobile products, personalization engines, and data-driven systems that solve real-world problems.
Education
UC Merced — Computer Science & Engineering
Focus
Mobile products, scoring engines, science-backed personalization
Current project
CurlCue — science-backed product intelligence for curly hair
Current status
Science-backed, profile-driven, and built like a real product.
CurlCue has a working scoring engine, 100+ classified ingredients, 5-step profile onboarding, role-aware routine management, and benchmark testing — with ongoing work on product coverage and TestFlight deployment.