ComingUp ComingUp
Appcroft – AI Try-On & Stylist for Shopify

Appcroft – AI Try-On & Stylist for Shopify

Shopify App Store-approved app: AI virtual try-on from one photo + AI stylist

Posted on behalf of the maker

Is this your project? Ask us to take it down

Gallery

Appcroft – AI Try-On & Stylist for Shopify

About

Ready to take over and launch. A finished Shopify app, approved and live on the Shopify App Store since 24 October 2026. Nothing is left to build: the new owner only needs to market it. No ads, content or outreach have been done yet, so every growth channel is open from day one.

App Store listing: https://apps.shopify.com/appcroft-ai-virtual-try-on

Demo video: https://vimeo.com/1242595554

What it does

  • Virtual try-on from one photo. A shopper uploads one full-body photo and in about a minute sees the store's clothes on themselves: a single piece or a whole outfit, placed in a scene (street, office, beach), even combined with clothes they already own. Then they add the outfit to the cart.
  • AI stylist / shopping assistant. Answers in the shopper's language, recommends only the store's own products with real sizes and stock, and opens the try-on straight from the chat.
  • Insights for the merchant. Try-ons, add-to-carts, orders and revenue, counted only when the cart line itself came from the try-on or the assistant.
  • Merchant controls. Image model, which products offer the try-on, scenes, wording and button style.
  • Private by design. Photos are deleted after 1 to 90 days, hosted in the EU, encrypted at rest, never used to train models.

Monetisation (already built on Shopify Billing)

  • Plans: Free, Starter $19, Growth $49 and Pro $149 per month (20% off yearly, 7-day trials)
  • One-time credit packs ($14 / $57 / $109) and automatic top-ups
  • Merchants can resell try-on credits to their own shoppers

Tech and operations

  • TypeScript monorepo with 2,175 unit and 265 integration tests
  • One Hetzner ARM server (Docker Compose behind Caddy), PostgreSQL + pgvector, Redis/BullMQ, S3-compatible storage, self-hosted embeddings
  • AI through fal.ai (Google Nano Banana, ByteDance Seedream and OpenAI GPT Image, the merchant picks one; BiRefNet for cut-outs) and OpenRouter for photo checks and the assistant
  • Shopify theme app extension plus a customer-account extension
  • Operator console for shops, money, incidents and the support mailbox
  • Encrypted data volume, encrypted off-site backups with weekly restore drills, scripted deploys and server moves
  • A few hours a week to run: merchant support and Telegram alerts. Backups, data retention and disk clean-up run automatically.

Status and costs

Pre-launch: approved and ready to launch, but not launched or marketed yet. 0 users, 0 customers, 0 revenue. Hosting runs at roughly $13 a month, the domain about $1 a month; AI costs scale with usage and are covered by the credits merchants buy. Target: Shopify fashion and apparel stores (clothing, shoes, accessories), worldwide.

Included in the sale

  • Complete source code with tests, documentation and deploy, backup and server-move scripts
  • The Shopify app and its approved App Store listing, with screenshots and copy (transferred to the buyer's Shopify Partner organisation)
  • The domain appcroft.com with the landing page, privacy policy, terms and the contact@ mailbox
  • A demo development store with a 368-product catalogue
  • Post-sale handover support from the founder

Risks to weigh

Demand is not proven yet (not launched); the virtual try-on category on the App Store is crowded; the app depends on Shopify and on AI providers and their prices (three image models are already supported, so the model layer is swappable); and the pinned MinIO storage image is no longer publicly downloadable. It runs fine and is backed up, but should be replaced by local disk or another S3 server before any future server move.

Comments (4)

Merlin Mosciski Merlin Mosciski 6 days ago

virtual try-on sounds cool, but how accurate is it really? does it actually help with styling choices?

Percy Weimann Percy Weimann 6 days ago

how does the AI handle different body types and skin tones in the try-on feature?

Camryn Barrows Camryn Barrows 6 days ago

another AI try-on app? seems like the market is flooded with these already.

Randall Ledner Randall Ledner 7 hours ago

looks promising, but I wonder how accurate the try-on really is.