Cartechelectronics

Designed and built the operational platform that Cartechelectronics runs its business on — the UK’s #1 provider of engine ECU services, offering repairs, cloning, genuine plug-and-play replacements and diagnostics, shipped worldwide.

Storefront, workshop and shipping run through one system rather than three, and the platform is under active development.

The problem

A business like this runs on physical objects moving through stages. A unit is posted in, it is diagnosed, repaired or cloned, then shipped back — often to another country. Meanwhile the storefront has taken an order, the workshop has a job, a carrier needs a label and customs paperwork, and stock has to reflect what is actually available to sell.

Those are separate systems with no shared model of an order, a job or a shipment. Left alone, the integration layer between them is a human — someone holding three interfaces open and reconciling them by hand, which is the most expensive possible way to run a business process.

Constraints

Every unautomated boundary between two systems is where the errors live: a wrong address on an international shipment, a job that does not match the order it belongs to, a customer asking where their part is and nobody able to answer quickly. Throughput is capped by how fast a person can move data across those boundaries, and accuracy degrades exactly when volume rises.

Worldwide shipping raises the cost of each of those mistakes, because the feedback loop is measured in days and the customer is usually without their vehicle for the duration.

What I built

Focusing end to end — architecture, build and ongoing development.

A Django system that owns the path between the storefront, the workshop and the carrier, so an order becomes a job and a job becomes a shipment as a single automated flow rather than three disconnected ones. Inventory is kept in step with Shopify, shipping and carrier paperwork are generated rather than typed, and ECU data extraction feeds the parts of the process that depend on knowing what a unit actually is.

The engineering interest is in the seams rather than in any one system: reconciling records that each service names differently, absorbing a third-party API being briefly unavailable without losing an order, and keeping the automation observable enough that staff can see what it did and step in when a case is genuinely unusual.

Outcome

The business runs on it. Orders, workshop jobs and international shipments move through one automated flow instead of three systems reconciled by hand — which lifts the ceiling on volume and removes the class of error that only surfaces once a package is already in another country.

This commerce integration work inspired the development of Shopify for GitHub — the same problem, getting store context in front of the people who have to act on it, generalised into a product.