Idol Fairies
A working K-pop merch e-commerce and finance automation ecosystem — built, automated, and tested end to end.
This is a portfolio/demo environment. Everything here — the storefront, orders, customers, resellers — runs on realistic simulated data, built specifically to demonstrate a production-grade system safely without exposing a live business.
70 / 70 regression tests passed across all 8 automation workflows (August 16, 2026) — Verified via live webhooks, real Gmail/Telegram triggers, and direct Supabase state checks — not just workflow execution status.
One connected system, not a pile of separate automations
Idol Fairies is a working K-pop merch e-commerce system, built end-to-end as a portfolio demonstration: a real storefront, real backend automation, and real reporting — running on synthetic order data so the mechanics can be shown safely without exposing a live business.
Two kinds of orders enter the same system: a shopper checking out on the storefront, and a reseller placing a wholesale order. From that point on, both flows share the same pipeline — sales records, inventory, receivables, dashboards, and the AI assistant all work off the same data, not separate systems bolted together.
Explore each part of the system
Every area below follows the same pattern: the business problem, what it does, what changes, and the proof — with the technical detail one click away for anyone who wants it.
Storefront & Ordering
Business problem
A K-pop merch store needs an online storefront where customers can browse, check stock accurately, and check out — without the owner manually tracking what's sold out.
What it does
A public storefront (shop, category pages, product pages, cart, checkout) shows only two real states — Available or Sold Out — pulled from the same inventory data the rest of the system uses, so there's no separate 'website copy' of stock that can drift out of sync.
What changes
No manual product/stock updates on a second system; what's shown to shoppers is always the same number used internally.
Built on Next.js 16 / React 19. Checkout hands off directly into the same order pipeline verified in the regression suite.
Want something like this for your business?
This is what a manual process looks like once it’s automated properly — tell me about yours.