Skip to content

Shopify · Supplier CSV

Her-age CSV_CMD — supplier dumps in, a quality gate, taxonomies out. The UI is theirs. The pipe is ours.

CSV_CMD is Her-age's internal tool: take supplier CSV from Shopify and from custom CMS / ERP, normalize, run an AI quality filter, apply taxonomies for import or export. Client designed the UI — we say so. Fastify on the back, Next.js dashboard, Remix + Polaris as the embedded Shopify app. File converter, auto-updater, pattern matcher, AI analyzer, CSV editor, taxonomies. Integration speed +56%. Fully automatic. Uptime over a year without intervention.

Her-age 2025 5 min read
Her-age case study
integration speed
+56%
end to end
Auto
no intervention
>1 yr

Many sources, one gate

Suppliers do not arrive in one shape. Shopify exports, custom CMS, ERP dumps. CSV_CMD ingests them, normalizes the rows, then a quality filter: no photo, three or more empty critical fields, duplicates, price and size anomalies. What survives gets a taxonomy — for import into the store or for export out.

Their screens, our pipe

The UI was designed by Her-age. We built the engine. Fastify backend. Next.js dashboard. Remix + Polaris for the embedded Shopify app. AWS and DigitalOcean, cron on the clock. GPT-2 sits in the listed stack as the AI analyzer — we do not invent model scores on top of that.

  • File converter
  • Auto-updater
  • Pattern matcher
  • AI analyzer (quality gate)
  • CSV editor
  • Taxonomies for import and export

The third Her-age product

Not the store redesign. Not Sell with Us. Not NFTIZE. CSV_CMD is the supplier integration tool that kept running. +56% faster to integrate. Automatic. More than a year up without someone in the seat.

Technologies

  • Fastify
  • Shopify
  • Remix.js
  • Next.js
  • GPT-2

Outcomes

  • +56% integration speed versus the previous supplier path
  • Fully automatic ingest → filter → taxonomy → import or export
  • Uptime over a year without intervention
All cases

Blog

The case for boring web stacks

TypeScript, Postgres and edge deployment are not exciting — they are how you still iterate fast in year three without rewriting the platform.

Blog

Shipping evals before prompts

Why we run regression suites on LLM outputs before every release — and how it changed our AI delivery cadence at ZNN.

Blog

Mobile release trains that do not break

Fastlane, staged rollouts and feature flags — a practical playbook from eighteen App Store submissions without a Friday-night rollback.

Luxury · Volume eDTL

The Brand Collector

The Brand Collector resells pre-owned luxury through a supplier network. Same class of problem as a catalogue feed — except each dump crossed 120,000 rows, and the CSV they hold sits above 169,000. Fastify stayed the orchestrator. The heavy compute moved to Go workers. About 60% faster on taxonomy, mapping and pattern-match at that volume.

iGaming · Platform

Casino operator · NDA

Under NDA we stabilised three production gambling brands: cleaned up legacy PHP/JS stacks, extracted shared modules, and shipped a new marketing site without stopping live traffic.

Marketing · Fintech

Digital DAO

Digital DAO is a marketing studio in Belarus that works mostly with fintech. We did not repaint the old site. We rebuilt the story and the visual system so it would sit in a room with banks, not with a carnival agency deck.