SkinNear

AI skin insights turned into a personalized skincare routine that fits your budget, with products you can find nearby.

Comment

Inspiration

Understanding your skin is only the beginning. You still have to choose products, work out whether the entire routine fits your budget, and find sellers in your country.

We built SkinNear to connect those decisions and make skin insights useful for people who feel overwhelmed by skincare choices or have a limited budget.

What it does

SkinNear turns a photo into cosmetic skin insights and a personalized skincare routine.

  • Simple capture: Take a photo using the camera or upload one, then crop, rotate or zoom before analysis.
  • Interactive insights: Explore YouCam skin scores and overlays through an original-image comparison gallery.
  • Care by Area: View observations for the whole face, T-zone and U-zone.
  • Voice explanations: With consent, Gemini explains the report aloud, accompanied by a transcript that follows playback.
  • Budget-aware routines: Recommendations account for country, currency, ingredient preferences and products users already own.
  • Budget Lab: Compare complete baskets and inspect product swaps before applying a choice.
  • Regional discovery: Explore 267 products across 26 brands, with seller searches and Blinkit and Zepto lookups for India.
  • Take the routine with you: Save products, download a shopping list or share a routine.

Essentials come first. Serums and retinol have separate consent and suitability checks, and any additional step must fit the remaining budget. Retailers confirm stock and final prices.

How we built it

We built SkinNear with Next.js, React and TypeScript, using reusable components and service classes. MongoDB supports persistence and shared production quotas, while Python supports catalog preparation.

The YouCam API integration handles file registration, signed uploads, asynchronous task creation and polling. We normalize scores, classifications and image metadata into a consistent report.

Product embeddings retrieve relevant candidates from the curated catalog. Deterministic checks then enforce eligibility and whole-routine budget constraints. Gemini explains the report and selected routine, while GSAP powers SVG animations and interface motion.

Challenges we ran into

Photo handling: The integration required signed uploads, asynchronous tasks and support for different response structures. Large phone photos, orientation and face framing needed careful handling and useful recovery guidance.

International shopping: Currency conversion does not establish local availability. We separated planning estimates, regional catalog eligibility and seller discovery.

Voice synchronization: Aligning the transcript with generated speech while preserving product names required more than a standard audio player.

Complete-basket budgeting: Individually affordable products can still create an unaffordable routine. We needed selection logic that checks the combined cost.

Accomplishments that we're proud of

We connected skin analysis to an actionable flow: understandable insights, product selection, visible budget trade-offs and regional shopping options.

Budget Lab lets users see what a different spending cap changes without another scan or paid AI explanation.

Camera capture, local cropping and upload fallback make getting started easier. Voice explanations and interactive overlays offer different ways to understand the results.

Additional skin observations do not automatically become additional purchases.

What we learned

Reliable API integration depends on preparing input, tracking tasks, interpreting results and helping users recover from errors.

Separating retrieval, selection and explanation gave us greater control: embeddings find candidates, explicit rules enforce constraints, and AI explains the outcome.

Personalization involves practical constraints as much as skin concerns. Budget, existing products and regional access determine whether a recommendation is useful.

What's next for SkinNear

We want to expand verified regional product coverage and improve the freshness of pricing and seller information. Direct retailer integrations would make availability more dependable.

We also plan to improve capture guidance, broaden language support for voice explanations and test across more devices and photo conditions.

Longer term, we want to help users revisit their routines and compare observations over time, with clear consent and privacy controls. SkinNear will remain focused on cosmetic education and practical shopping support.

Built With

Share this project:

Updates

Submission history