Inspiration

Every major milestone, from weddings to keynotes, comes with months of wardrobe planning and photography prep, yet skin prep is almost always treated as an afterthought. People routinely start aggressive treatments days before an event, resulting in compromised skin barriers, irritation, and unexpected breakouts right before camera day. At the same time, standard virtual try-on tools treat clothing in total isolation, ignoring how fabric tones, sheen, and studio lighting interact with a person’s actual skin finish. We built Ready to solve this disconnect by orchestrating skin health and wardrobe styling into a unified, biologically timed 42-day event runway.

What it does

Ready synchronizes clinical-grade skin analysis with virtual apparel styling across a structured countdown. Using Perfect Corp’s YouCam Skin AI, the app analyzes eight core skin metrics, including hydration, oil balance, redness, and texture, from a single photo. A dedicated Gemini agent acts as both a cosmetic chemist and stylist, converting these diagnostics into a safe weekly regimen. To protect the user's barrier, the system enforces a strict lockout on new active ingredients during the final seven days before an event. Alongside the regimen, users test curated outfits while receiving tailored grooming and makeup finish recommendations, such as pairing luminous fabrics with matte skin bases. Progress scans at key milestones dynamically update the routine as the event approaches.

How we built it

The stack pairs Next.js 16, TypeScript, and Tailwind CSS on the frontend with server-to-server integrations across external APIs. We implemented backend handlers for Perfect Corp’s YouCam skin and cloth services, managing secure presigned asset uploads, asymmetric RSA key authentication, and polling workflows. Google Gemini powers the decision engine, reasoning through skin metrics to output structured calendar milestones and ingredient safety rules. To keep the app responsive and cost-efficient, we added a caching layer that eliminates redundant analysis calls when processing repeat states.

Challenges we ran into

Building the integration presented several technical hurdles. YouCam’s presigned upload endpoints strictly validated MIME types, causing silent 403 errors until we built server-side extraction to dynamically parse and forward exact file headers. We also had to solve a fundamental UI framing conflict: skin diagnostics require a high-resolution close-up portrait, while apparel try-ons demand a full-body silhouette. We resolved this by creating an architectural silhouette canvas that dynamically matches calibrated skin tones to seamlessly bridge headshot scans with full-outfit previews. Finally, we implemented fallback layers to gracefully handle third-party model deprecations and API rate limits.

Accomplishments that we're proud of

We are particularly proud of bridging the gap between clinical skin metrics and aesthetic styling, creating an engine that directly links skin barrier needs to wardrobe fabric pairings. The entire build was shipped with zero credential exposure across our repositories, complete with sub-second client interactions and automated fallbacks that prevent broken states during evaluation.

What we learned

The process reinforced that skin turnover cycles cannot be rushed; an effective prep tool must respect human biology rather than promise overnight fixes. We also gained extensive experience architecting secure, enterprise-grade server-to-server pipelines with asymmetric encryption and asynchronous job polling.

What's next for Ready

Looking ahead, we plan to connect Ready directly to beauty and fashion retail APIs for one-click checkout of curated routines and garments. We also aim to incorporate real-time environmental data, such as local humidity and UV indices, to auto-adjust regimens leading up to event day, alongside shared group dashboards for wedding parties and speaking panels.