StylePilot Portfolio

StylePilot
Your AI Personal Stylist
  Judy Xiong, Founder/CEO/Product Lead
Overview
StylePilot is an AI-powered personal styling copilot that helps people feel confident in what they wear—using what they already own. The app digitizes a user’s wardrobe and delivers personalized, context-aware outfit suggestions based on weather, events, body profile, and style preferences, eliminating the daily “nothing to wear” problem without relying on constant shopping.
By combining computer vision, large language models, and contextual recommendation systems, StylePilot transforms a fragmented closet into an intelligent styling experience: lowering the cost of personal styling, reducing overconsumption, and helping users get more value from their existing wardrobe.
StylePilot
Style Made Simple.
Product Demo
Loading...
Customer Challenge
Closet full but still feel NOTHING TO WEAR
Busy professionals and working moms fall into DECISION FATIGUE about what to wear every morning
Shopping fatigue: BUY NEW instead of reusing
Stylists cost $200–500/hour → INACCESSIBLE
Gen Z & Millennials want SUSTAINABILITY

AI Hypothesis
“If I deliver the AI result in the form of outfit suggestions, virtual try-on previews, and wardrobe optimization with ≥80% tagging accuracy and fast (≤3s) response time to busy professionals and working moms, they will be able to reduce daily decision fatigue, reuse more of their wardrobe, and shop more confidently, which in turn creates higher confidence, saved time, and reduced shopping waste — and I can capture value through a $9.99/month premium subscription and affiliate commissions on smart shopping suggestions.”
Customer Interview
Persona 1 - Busy Professionals
Bryan, Executive Manager (Age 35)
Lifestyle
Lives in a busy city, has a closet full of clothes but limited time to decide what to wear.
What He Struggles With:
●“Too many clothes but nothing to wear.”
●Outfit combinations are difficult.
●Styling feels repetitive.
What Bryan Wants:
●Look stylish and confident at work and social events.
●Save time choosing outfits.
●Reuse existing wardrobe pieces smartly.
What She Does:
●Takes photos of outfits for OOTD (Outfit of the Day) and inspo.
●Shops based on style, trends, and price.
Strong interest in sustainability and reusing outfits.

Persona 2 - Working Moms
Emily, Working Mom (Age 38)
Lifestyle
Lives in a suburban or urban area, balancing a full-time job with raising young children. Her days are packed with work meetings, school drop-offs, errands, and family activities, leaving little mental space for outfit planning.
What She Struggles With:
“I want to look put-together, but I’m always short on time.”
Clothes don’t feel practical and stylish at the same time.
Repeating the same outfits because trying new combinations feels exhausting.
What Emily Wants:
Look polished and confident at work, school events, and social outings.
Spend minimal time thinking about what to wear.
Make better use of her existing wardrobe without buying more.
What She Does:
Gets dressed in a rush, often defaulting to “safe” outfits.
Values convenience first, sustainability second.
Uses her phone frequently for planning, reminders, and daily decisions.

User Journey Map

MVP Feature Priority Matrix
MVP Solutions
1
Smart Wardrobe Digitization
Scan and digitize your entire closet with computer vision technology
This feature enables users to digitize their physical wardrobe by uploading or taking photos of their clothes. The AI system automatically detects and categorizes clothing items, creating a digital wardrobe foundation for styling, sustainability tracking, and recommendation features.

2
AI Auto-Tagging
Automatic classification by color, garment type, style preferences, and occasion suitability.
The AI system automatically detects and categorizes clothing items, creating a digital wardrobe foundation for styling, sustainability tracking, and recommendation features.
3
Contextual Outfit Generator (OOTD Recommendation + Mode-based Outfit Planning)
Personalized outfit recommendation based on event type, weather, and your unique style profile
The Homepage integrates schedule, weather, and styling tips for quick, scenario-matched outfit solutions. Users can select occasions (e.g., Trip, Party) and input specific details for tailored recommendations.
4
RAG-Based Stylist Chatbot
Conversational AI stylist providing real-time fashion advice and outfit refinement
The Chatbot is an interactive assistant that delivers personalized outfit recommendations via two core sub-features: "Surprise Me" and "Recreate Outfit", enabling users to quickly get tailored styling ideas.
5
2D Virtual Try-On
Visualize outfits before really getting dressed
The User Photo Upload feature allows users to provide a clear, full-body photo—either from their device gallery or by taking a new picture—to enable the Virtual Try-On experience. Uploaded photos will be used to generate personalized outfit visualizations.
6
Sustainability Tracker
Monitor wardrobe usage frequency and gain AI insights into wearing patterns; provide restyling tips to forgotten pieces for over 60 days
The Sustainability Tracker Feature enables users to record, track, and analyze the sustainable attributes of clothing in their digital wardrobe, covering the entire lifecycle of garments from production to disposal.
MVP Metrics
North Star Metrics:
Number of outfits worn from existing wardrobe (where the outfits suggested is new to the users)
Number if outfits adopted by user (per week)

Signups
No. of users sign up to StylePilot

Conversion Rate: No. of users converted from freemium to paid
(Optional) Days it took to convert users from freemium to paid

Activation
Time to onboard user to StylePilot
Time to upload images
Time taken to generate (first set) of outfits
Average time taken to generate subsequent outfits

Engagement
Daily Active users
Average outfit generation per week

Retention
Measurement of Daily active users MoM (month over month)
Number of dropped users bi weekly
Reactivation rate (monthly)

Revenue
Free to premium conversion
Items purchased vs use of Outfits generated
Carbon savings with the adoption of every outfit suggested.

AI Model Data Inputs and Outputs
AI Model Deep Dive
AI Model Selection
Computer Vision Tagging
Summary
Across the evaluated vision tagging models, Gemini 2.5 Flash Lite delivers the strongest overall performance, offering the best balance of speed, cost, and tag coverage. It achieves near-top tag extraction (786 tags) with low latency (2.68s) at an order-of-magnitude lower cost ($0.0066) than alternatives. Gemini 3 slightly outperforms on tag volume and latency but at significantly higher cost, making it better suited for accuracy-critical use cases. In contrast, GPT-4o Mini and Claude Sonnet 4 underperform for large-scale vision tagging, with slower response times, fewer extracted tags, and substantially higher costs, resulting in lower overall ROI for production tagging workloads.
AI Model Selection 
Vision Tagging - Gemini 2.5 Flash vs Gemini 3

AI Model Selection
 Chat Copilot
Summary
GPT-4o was selected as the primary model because it offers the best balance of speed, consistency, and reliability for real-time consumer UX.
While Claude slightly outperforms on deep reasoning, its latency (~5× slower) makes it unsuitable for interactive styling flows. Gemini showed high variance and critical failures, which undermines user trust.
For a conversational AI stylist, fast, stable, and predictable responses matter more than marginal reasoning gains—making GPT-4o the best production choice.
System Architecture (As-is)
To-be Architecture (Scalable Future)
System Data Flow 

Product Roadmap

Value Proposition

Go-to-Market
GTM Strategies

North Star Metrics

Business Models
Why
StylePilot is more than a styling app—it's the emotional bridge between how we live, how AI empowers us, and how we move toward a more sustainable future.
Our Mission: We change how you feel about yourself from day to day by transforming your relationship with your wardrobe.
Core Features Demo
Digital Wardrobe and
OOTD Recommendation 
Outfit of the day recommendation based on calendar, event and style preferences.

Mode-based Planning
Planning the outfit based on different occassions such as trip, work, gym, event, etc. 
AI Chatbot and Styling Tips
Your AI Personal Stylist, 24 hours/7 days


2D- Virtual Try-on
Virtual try-on fit into different occassions. 

Sustainbility Tracker and
Rewear Intelligence
Keep track of the forgotten outfits over 90 days and provide restyling tips to increase the utilization of existing wardrobe. 


Contact us
Contact: team.stylepilot@gmail.com
Website: https://stylepilot.ai/
Beta Testing Link: 
Click here