Taewon Kwon

Zipper distributor

Bringing a Korean zipper distributor founded in 1970 to a younger generation of buyers, and getting 4 million units of inventory off paper.

Role
Sole builder and designer, end to end
Type
Paid client engagement
When
2026 [TK: start month] – Oct 2026
Status
Delivered Oct 2026
Stack
Flutter (Android kiosk), web dashboard, Supabase
Scope
Inventory system · website & SEO · sample book · cards

A long-established distributor of zippers and apparel materials ran on decades-old relationships and paper stock records. I delivered four things under one fixed deadline and budget: its first digital inventory system, a website with search and maps presence, a redesigned sample book, and business cards.

4M
units in stock
20,000
SKUs
~10,000
stock movements a week, previously tracked by hand

The problem

The business ran on relationships with older client companies. Those still worked, but younger buyers find suppliers through search and maps, and this company barely showed up in either.

Inside, inventory lived on paper. Around ten thousand stock movements a week were recorded by hand across twenty thousand SKUs, so knowing what was actually on the shelves depended on people and memory.

Train of thought

The filled markers are where the plan changed.

  1. Scoping [TK: month]Two goals that look unrelated, to be found by new buyers and to know what’s in stock, share one constraint: a fixed budget and a team that doesn’t use software day to day.
  2. On the warehouse floorThe people recording stock movements are warehouse staff, not office users. If the system needed training, it wouldn’t be used. That led to a kiosk with four buttons, and to an append-only ledger behind it so nothing could be silently overwritten.
  3. Website scopeThe tempting version had a content pipeline and a blog. What actually moves search and maps is being findable at all. I launched with maps listings and base SEO, and deferred the blog.
  4. Budget pressureManufacturing the sample book took most of the physical budget. Rather than cut quality, I designed every page and the binder myself, and the manufacturer recreated it from my files. The printed catalogue was scrapped, and the physical rebrand narrowed to cards and the sample book.
  5. Planning the cutoverMoving a live business off paper risks a week of wrong numbers. I planned an order freeze and a full data sweep, and I committed to staying until the system had run live for at least a week.
  6. Sep – Oct 2026Site live in mid-September, inventory cut over from paper in September, and all four deliverables handed over by October.

Inventory system

The company’s first digital inventory system, in three parts: a Korean-language kiosk app on an Android tablet on the warehouse floor, a web dashboard for the office, and a Supabase backend.

KioskAndroid tablet on the warehouse floor, Korean UI, four buttons
Append-only ledgerEvery movement is a new row; nothing is edited in place
Office dashboardStock levels and history for the office
One source of truth. Staff record movements, the ledger keeps every one, and the office reads stock levels from that history.
Button 1
[TK]
Button 2
[TK]
Button 3
[TK]
Button 4
[TK]
TimeSKUMovementQty
09:12ZP-0000In+500
09:40ZP-0000Out−120
Illustrative rows, not real data [TK: real button labels and ledger columns]
Four buttons, every movement kept. Because the ledger is append-only, every stock movement stays on record and auditable, without asking staff to learn software.
Image: the kiosk on the warehouse floorTablet in place, with staff if they agree to it [TK]
Where it lives. [TK: caption.]

The cutover

Switching a live business from paper to software is where these projects usually fail. Orders keep moving while you migrate, and the first week of numbers decides whether anyone trusts the system.

  1. Friday afternoonOrders frozenNo movements while the data is swept
  2. SaturdayFull data sweepStock recorded into the new system
  3. First week liveRun, watch, fixErrors surface while I can still fix them
  4. Sep 2026Paper retiredInventory cut over from paper
A controlled cutover. I required at least a week of live running before leaving, so problems showed up while I could still fix them.

Search presence

A new site, email migration, Google SEO and Naver Place registration, live since mid-September. I deliberately stopped at the basics: maps and base SEO first, with the blog pipeline deferred until the basics were working.

Image: the new websiteDesktop screenshot; check whether it shows the company name [TK]
The new site. [TK: caption]

Sample book

Two binders that show the physical zipper samples by product line. Manufacturing took most of the budget, so I designed every page and the binder myself, and the manufacturer recreated it from my files.

  1. Five directions.I showed the client team five different designs.
  2. The team picked on taste.They know their buyers, so the choice was theirs.
  3. The manufacturer checked feasibility.Each direction was tested against what could actually be produced.
  4. One final version.It combined the strongest parts of all five.
Image: the five directions side by sideSpreads or mockups [TK]
Five directions. [TK: caption]
Image: the finished binderPhoto of a spread with real samples [TK]
The final book. [TK: caption]

Outcome

All four deliverables were handed over by October 2026, with the inventory system cut over from paper in September and the website live since mid-September.

[TK: anything measurable since handover: search or map inquiries, how the floor staff took to the kiosk, the first stock count against the ledger.]

What it shows: end-to-end delivery for a non-technical, traditional business, across software and physical work, under a fixed deadline and budget.