Inspiration
Online fashion shopping gives us endless options, but finding clothing that matches personal coverage preferences can still feel surprisingly difficult. Shoppers can filter by size, colour, price, and category, yet details like neckline, sleeve length, opacity, garment length, slits, or open backs often require manually checking each item.
We also wanted to move away from the idea that modesty has one universal definition. Modesty is a personal journey, and what works for one woman, or even for the same woman in different situations, can vary.
Inspired by TechNisa's Broken Systems theme, we asked:
What if fashion search adapted to the shopper instead of making the shopper adapt to it?
That became Her Closet.
What it does
Her Closet is a personalized modest-fashion discovery platform that brings products from multiple retailers into one unified shopping experience.
Users can create and save their own coverage preferences, shop for different occasions, and browse products filtered around what feels right for them.
Her Closet includes:
- Personalized modesty profiles
- Occasion-based shopping
- Interactive visual onboarding
- Smart coverage filtering
- Temporary session adjustments
- A 7-point AI Modesty Audit
- A multi-store Closet Hamper
- An estimated shopping receipt
- Direct links back to original retailers
The AI Audit breaks garments down by characteristics such as neckline, sleeve length, opacity, slits, cutouts, hemline, and fit, helping users understand why an item received its result.
Most importantly, Her Closet does not decide what modesty means for the user — it gives them the tools to decide for themselves.
How we built it
We built Her Closet using:
Next.js, React, TypeScript, Tailwind CSS, AI Vision / Computer Vision, Custom Modesty Scoring Logic, Custom Filtering Engine, Multi-Retailer Product Data, Urban Planet, Ardene, PixelLab AI, Git, GitHub, npm, Canva, TypeScript Compiler, Next.js Production Build, Antigravity, Gemini, ChatGPT
Products from retailers such as Urban Planet and Ardene are organized into a shared catalog so users can browse multiple stores from one feed.
Our filtering engine compares product attributes with the shopper's saved preferences and temporary session filters.
For the AI Audit, garment characteristics are analyzed and organized into a 7-point breakdown so shoppers can understand why an item received its result instead of only seeing a score.
We connected the entire shopping journey:
Profile → Personalized Feed → AI Audit → Hamper → Estimated Receipt → Retailer
Challenges we ran into
One of our biggest challenges was realizing that clothing rules cannot be applied equally across all garment categories.
For example, pants should not be penalized or fail a filter simply because they lack sleeves or a neckline, while cropped tops and asymmetrical cutouts require distinct parsing compared to standard-length garments.
We had to continuously refine our filtering and classification logic to correctly distinguish:
Tops and blouses
Pants and jeans (ensuring denim tops and jackets didn't bleed into bottom wear)
Skirts and dresses
Sweaters, hoodies, and outerwear
Cropped silhouettes and varied hemline lengths Asymmetrical cuts and differing silhouette fits Building this deterministic filter engine was a major technical hurdle. Mainstream retail tags are notoriously inconsistent, a cropped knit tank might be loosely tagged as a "sweater," or a short-sleeve crewneck might be mislabeled under generic top filters. Establishing fail-safe keyword whitelists and regex boundary checks to prevent exposure leaks across thousands of catalog items proved complex, and fine-tuning these edge cases remains an active work in progress.
Another key challenge was balancing personalization with judgment.
We did not want hercloset to impose a single, rigid definition of modesty. This led us to decouple garment analysis from user preference: our computer vision and classification pipeline identifies concrete clothing attributes, while each shopper defines what coverage works for them.
Finally, orchestrating catalog aggregation, AI audit models, dynamic state calculations, and a cohesive UI within a tight hackathon timeline required constant prioritization and rapid debugging.
Accomplishments that we're proud of
We're proud that Her Closet became more than a simple filter page, we built an end-to-end shopping experience.
Some of the features we are especially proud of include:
- Unified multi-store fashion discovery
- Personalized coverage preferences
- Occasion-based shopping
- Interactive pixel-art onboarding
- Smart coverage filtering
- A detailed 7-point AI garment audit
- Live modest-match feedback
- A Pinterest-inspired product feed
- A multi-store Closet Hamper
- An estimated shopping receipt with retailer links
We're especially proud that the product stayed true to one principle throughout development:
Personalization without prescribing what modesty should look like.
What we learned
We learned that building a useful AI product is not only about generating an answer, it is also about deciding what AI should and should not decide.
For Her Closet, AI is most useful for identifying and explaining garment characteristics. The final decision remains with the shopper.
We also learned more about:
- Filtering systems
- Product classification
- Working with multi-retailer data
- TypeScript
- React and Next.js
- Component-based UI design
- Git and GitHub collaboration
- AI-assisted garment analysis
- Building and debugging quickly under a deadline
Most importantly, we learned how much stronger a product becomes when technology, design, and purpose all support the same idea.
What's next for Her Closet
We would love to expand Her Closet by adding more retailers and a much larger fashion catalog.
Future versions could include:
- Improved computer-vision garment analysis
- More detailed coverage preferences
- Personalized recommendations
- Saved outfits and collections
- Cross-store outfit building
- Clothing uploads
- Layering suggestions
- More occasion-based customization
- Smarter personalization over time
Ultimately, we want Her Closet to become a personalized discovery layer across fashion retailers, helping shoppers spend less time searching and more time finding clothing that works for them.
Log in or sign up for Devpost to join the conversation.