Inspiration
Fashion and cosmetic preparation are usually treated as separate decisions, even though they shape the same moment. We wanted to replace disconnected product browsing with one clear journey built around where the shopper is going and how they want to feel.
What it does
Shoppers choose an occasion, style, formality, and budget, then provide consented face and full-body photos. MirrorMoment recommends three compatible garments, generates virtual try-ons through YouCam AI Clothes V3, adds optional non-medical cosmetic context from YouCam Skin Analysis, and creates a downloadable final plan.
How we built it
We built MirrorMoment with Next.js, React, TypeScript, Tailwind CSS, and server-only YouCam integrations. The app handles YouCam’s asynchronous upload → task → polling → result workflow, preserves successful results during partial failures, and stores only limited session data in the browser.
Challenges we ran into
The main challenges were handling several independent asynchronous tasks, keeping completed looks available when another failed, and satisfying YouCam’s image-framing requirements. We also had to protect the API key, normalize raw Skin Analysis results, and distinguish recorded replay results from fresh live API calls.
Accomplishments that we're proud of
We completed a live end-to-end run with one Skin Analysis task and three distinct Apparel VTO results. We also created a transparent key-free judge replay, individual look retries, session controls, and a downloadable Confidence Plan—turning the APIs into a coherent product instead of a single-call wrapper.
What we learned
We learned how strongly input framing affects visual-AI reliability and how important visible progress and partial recovery are for asynchronous experiences. We also learned to communicate cosmetic insights carefully without making medical, sizing, or fit claims.
What's next for MirrorMoment
Next, we would connect MirrorMoment to real retailer catalogs and product pages, expand its occasion and cosmetic-personalization options, add comparison feedback from real shoppers, and strengthen production consent and vendor-data lifecycle controls.
Built With
- api
- next.js
- node.js
- playwright
- react
- tailwind
- typescript
- vitest
- youcam
- youcam-skin-analysis
- zod
Log in or sign up for Devpost to join the conversation.