Love the look. Know the details. Shop with your everyday needs in mind.
Set your essentials and preferences—from machine washing to functional pockets.
See what the listing supports, inspect the source, and find out what still needs an answer.
Compare your options against the requirements that matter to you.
Choose what should stay unchanged before your virtual try-on.
Review a protected-area change in our clearly labelled glasses-change demonstration.
A controlled repair example shows the glasses restored and the review updated.
Keep a record of your appearance review, then delete your session images.
Compare what arrived with the listing and record your own observations.
Help the next shopper with useful product observations submitted for moderation.
Manage your saved information, consent, exports, and deletion in one place.

Inspiration

A jacket can look perfect online and still be wrong for everyday life. Can it go in the washing machine? Are the pockets functional? Does it have the sleeves or closures you need?

Product photos rarely answer every question. Virtual try-on helps us imagine an outfit, but it introduces another question: did the preview change something about us along with the clothes?

We built AsShown to connect those decisions: understand the product, preview the look, and keep track of what you actually know.

Love the look. Know the details.

What it does

AsShown is a clothing-shopping companion that follows the journey from finding an item to checking it after delivery.

Start with your everyday needs. Add a product description and photos, confirm the color and size, and choose essentials and preferences such as machine washing, long sleeves, or functional pockets. Type or dictate your needs, with optional AI assistance to organize them.

See the evidence behind each finding. AsShown connects requirements to quoted source statements and keeps missing information visible. When a detail is unclear, it helps you draft a seller question and add the reply as further evidence.

Compare your options. Keep a shortlist, revisit saved checks, and reuse your requirements across items.

Preview with YouCam. Bring the garment into the studio, review photo-sharing consent, and define an Identity Contract covering appearance details you want protected, including glasses, face, hair, skin appearance, and background. Compare the result with the original, review flagged changes, and retain an appearance-review receipt. Optional AI garment observations support manual inspection; they do not establish fit or product quality.

Continue after delivery. Record how the item compares with its listing and contribute useful observations through a moderated community flow.

Privacy controls bring consent, exports, retention, and deletion into the same experience.

How we built it

We built AsShown with Next.js, React, and TypeScript, with server-side integrations for YouCam and OpenAI.

YouCam powers virtual clothing try-on. The studio manages image submission, generation status, comparison, and the review workflow around the result.

OpenAI supports optional assistance: organizing requirements, translating summaries, extracting text from evidence, and suggesting visible garment observations. Users review those suggestions before relying on them.

Our evidence system keeps each source attached to its product variant. A statement about one color or size cannot silently become evidence for another. Findings retain their supporting quotes, while missing information remains unknown.

We use Clerk for authentication, Neon Postgres for persistent records, Vercel Blob for private evidence storage, and Vercel for deployment.

AsShown extends our earlier KeepMe foundation into a broader shopping journey, connecting appearance preservation with product requirements, arrival checks, and community knowledge.

Challenges we ran into

The hardest challenge was deciding what the product could responsibly conclude. A photo can show a pocket without establishing whether it works. A generated preview can show a look without confirming comfort, washing instructions, or fit. We designed the experience to keep those distinctions clear.

Retailer imports were unreliable, so we removed the importer and focused on descriptions and photos supplied by the shopper.

Generative APIs also require careful recovery. An uncertain submission should not trigger blind retries that might create duplicate requests. We added explicit status handling and credit checks.

Finally, we worked to make a detailed evidence workflow approachable through short labels, progressive disclosure, and a consistent visual design.

Accomplishments that we're proud of

We connected requirements, evidence, seller questions, shortlisting, virtual try-on, arrival checks, and community contributions in one application.

We are especially proud that uncertainty has a useful next step. An unanswered question can become a seller reply, a recorded observation, or a clearer decision.

The studio makes appearance preservation visible and reviewable. Our demo includes a clearly labelled controlled glasses-change and repair sequence, distinct from live YouCam generation.

Keyboard navigation, dictation, read-aloud support, and optional translation help more people work through the same decisions.

What we learned

A useful shopping assistant needs to explain its conclusions and make unanswered questions easy to act on.

Product knowledge grows over time. The listing, a seller’s reply, and the shopper’s experience after delivery each contribute different evidence. Keeping their origins visible makes that information more useful.

Virtual try-on works best within this larger journey: exploring appearance while grounding practical requirements in product evidence.

What's next for AsShown

We want to refine AsShown with shoppers whose clothing choices depend on practical details, including people with accessibility needs and caregivers shopping for others.

Our priorities are broader evaluation of appearance checks, expanded language support, stronger community moderation, and reliable retailer integrations through supported APIs.

We also plan to connect confirmed care instructions with observations about washing and wear.

Our goal is simple: help people choose clothes they can enjoy, use, and keep.