All projects11 of 13 · Knitwear
11LIVE

IsoCode Knitwear

Order Tracking That Warns Before a Shipment Is Late

  • Web & Backend
  • Cloud & DevOps
  • AI / ML
  • Languages

One order falls behind, and says why

  • On time

  • On time

  • Behindneed 467 a day, making 350 a day

  • On time

  • No updatesgrey, never green

A picture of the idea. The rows are made up; the sum is the project's own example.

Each row is an order being knitted toward its ship date at the right edge. The upright line is where a row has to be today. One order slows, falls behind the line and turns amber well before the date, with the sum that says why. An order with no updates stays grey, never green.

The problem and the idea

Order tracking for a knitwear exporter that says, before it is too late, which orders will miss their ship date, and why. Supervisors send a daily count from a phone, in Tamil, even without a signal. A server works out for every order and stage whether it will ship on time and shows the reasoning. Built on demo data.

What I built

  • Every status colour comes with the sum behind it, for example: need 467 a day, making 350 a day
  • Supervisor screens are Tamil first and work offline: updates are saved on the phone and sent when it can, never counted twice
  • An order with no updates is grey, never green, and a rate based on little data is labelled an estimate
  • Buyers get a read-only page of only their own orders, and only what a merchandiser has confirmed
  • A supervisor's Tamil note gets an English meaning and a suggested reason with its confidence; the merchandiser decides
  • Row-level security is forced on every table

Will it ship on time?

An illustration of the idea, with numbers you set. It is not the product's own engine.

467
350
Daily counts in from the floor
Behindneed 467 a day, making 350 a day117 a day short.

Stack

  • TypeScript
  • React 19
  • PostgreSQL 16
  • Row-level security
  • GitHub Actions