When e-commerce isn't the problem. Product fit is.
Jaqueting sells custom jackets: the client configures the garment piece by piece, and the item is manufactured for them. That sounds simple until you put the system in production and find the configurator is just 20% of the real technical problem.
Jaqueting needed more than a site.
Jaqueting operates in a market where each jacket is unique: combinations of fabrics, linings, buttons, embroidery, custom sizes. The promise to the customer is that what they see in the configurator is exactly what arrives at their door in X days.
When they contacted us, they had a site that worked "sort of": the configurator built the garment, the cart took orders, but the system lived disconnected from production, inventory, and logistics. Each order required someone to manually "translate" it into a factory order.
What we found in discovery.
We spent two weeks mapping the real flow, not the documented one. The difference was key: the documented flow said each order traveled cleanly through six systems, but the real flow showed four of those systems were resyncing by hand via email.
The problem wasn't the configurator's frontend. The problem was that the piece couldn't be built without a technical backbone connecting configurator → production → fabric inventory → logistics → end customer.
Final architecture.
Next.js with real-time configurator, 3D render of the garment under construction, cart with automatic configuration save.
NestJS on PostgreSQL. API consumed by the configurator and exposing endpoints for the production system.
Webhook from backend to the factory system with automatic translation of configuration → cut orders → sew orders → packing orders.
Sync with fabric stock. The configurator disables options when a specific fabric is out, preventing the sale of what can't be produced.
Courier integration for automatic tracking and customer notifications.
Next.js, NestJS, PostgreSQL, Redis, Cloudflare Workers, Stripe.
Production metrics.
PROVISIONALReplace with verifiable real metrics before publishing.
“FastNet built the technical backbone I no longer worry about. I used to spend 10 hours a week checking if the systems were talking to each other. Now I spend zero.”