Solo build · One Piece TCG collection tracker

MyTCG

A collection tracker I designed and built end to end — card scanning, portfolio tracking, market pricing and wishlists across 2,255 cards. Built because I wanted one and none of them were any good.

0
Cards in the database
0
Variant records
0
Scan modes
0
Pricing regions — UK, EU, US
Role
Everything — research, IA, design, build, deployment
Team
Solo
Stack
Vercel · Neon Postgres · Clerk Auth
01 — The problem

A spreadsheet, a pricing app, and manual lookups.

The One Piece TCG community has no purpose-built collector tool. Existing solutions are fragmented, and collectors juggle several sources to answer basic questions — what is my collection worth?, is now a good time to buy this?

Price data is scattered across regional marketplaces with no single source.

Card identification is slow — looking up set codes and matching variants and printings by hand. The same card can have five or more printings at wildly different prices.

02 — The principles

Data density without clutter.

Editorial, not gamer

A minimalist editorial aesthetic taken from Wise and Bloomberg rather than from other card apps. Financial products already solved the problem of showing a lot of numbers calmly.

Accessibility first, personally

As a dyslexic user I enforced a 14px minimum body size, high-contrast type and scannable layouts. If it does not work for me it does not ship.

Progressive disclosure

A collection tracker can dump everything on screen at once. This one answers the question you asked and keeps the rest one tap away.

03 — What it does
Home dashboard

Portfolio value on an interactive chart, most valuable cards, recently added, price movers, and deals sitting below their 30-day average.

Search & browse

2,255 cards behind a set pill carousel, a slide-out filter panel and infinite scroll — with owned indicators and set completion bars.

Card detail

Variant-aware throughout — the same card can have five printings at wildly different prices, so the variant is never an afterthought.

News feed

AI-summarised community news in a newspaper-style grid, so the market context sits beside the portfolio rather than in another tab.

Three scan modes, because "scan a card" means three things

Price check

Scan, see the price, done. No commitment, no collection change.

Add single

Scan, select the variant, choose the collection. Variant selection is the critical step — the wrong printing is the wrong price.

Batch scan

Scan a stack, stage locally with a running summary, add them all at once.

04 — The calls

Removed the batch scan "Clear" button

Instead of

Keeping it and adding a confirmation dialog.

Why

I wiped 10 staged scans by accident during testing. Destructive actions should not sit next to routine controls — a dialog protects the action, it does not fix the placement.

Cost

Starting a batch over now takes more steps than it did. That is the correct trade for a stack of forty cards.

Built my own card database rather than calling external APIs at runtime

Instead of

Querying three third-party sources live, per request.

Why

A reconciliation pipeline merges the three into one Postgres schema with variant-aware records. Every API in the app reads from the database, so the interface is fast and the data is consistent.

Cost

A pipeline to maintain, and a reconciliation step that has to keep working as the sources change shape.

Region-aware pricing rather than one currency

Instead of

A single price in a single currency, converted at display time.

Why

Collectors operate in different markets. UK, EU and US each get their own currency and their own marketplace links throughout the app — a converted US price is not the price you can actually pay.

Cost

Three price sets to keep current instead of one, and every surface in the app has to respect the setting.

05 — Where it is

Live, and actually used.

A full end-to-end product: OCR scanning pipeline, AI-powered news feed, interactive portfolio charts, region-aware pricing, and QR-code collection sharing that generates public pages viewable without an account. Designed and built solo — from research and IA through to production deployment.

It is also the clearest answer I have to the question of whether a designer who writes production code designs differently. I never once specified something I could not cost, because I was the one paying for it.

Next
Cardconomy →

The venture. 92 screens, and one conversation that took the strategy apart.

declanamalone@gmail.com