Screenshots are not available for this project as it involves internal operational tools.
Goldelucks sends beautifully packaged gifts - donuts, brownies, cakes, cupcakes - to doorsteps across Melbourne and Sydney every single day. The product is considered. The packaging is meticulous. The operations, when we first spoke to them, were not.
The Process That Was Holding Them Back
Every night, someone on the team would open Shopify, manually copy the day's orders into a spreadsheet, and work through a series of calculations: what needs to be baked, when, what add-ons go in which box, who's delivering to which suburb. It was done overnight so the kitchen had a run sheet by morning.
The problems were predictable. Orders dropped off the list. Handoffs were verbal. If someone was sick, institutional knowledge walked out the door. High-priority same-day orders - the ones placed at 2pm for a 5pm delivery - sometimes didn't make it onto the sheet at all.
They weren't looking for a technology transformation. They were looking for the chaos to stop.
What We Built
We built a Shopify embedded app that sits directly inside their Shopify admin - no separate login, no context switching.
Webhook-driven order ingestion: Every order that hits Shopify is automatically ingested, parsed, and classified. The system reads their existing order tags to determine delivery location (Melbourne or Sydney), extract the delivery date, identify the products and add-ons, and flag urgency. Orders placed same-day after 12pm are automatically marked HIGH priority. No human judgement required for the classification.
Delivery dashboards: Two live dashboards - one per city - showing every order for the selected date. Ops can assign a driver to an order, mark it complete, and adjust urgency in a few clicks. The view is always current: changes made by anyone on the team appear instantly for everyone else via WebSocket, no page refresh required.
Daily Tally: Each morning, the kitchen opens a production tally that shows exactly what needs to be made that day - broken down by product type (donuts, cakes, cookies, loaded brownies, cupcakes) and then by variant. The tally is print-ready. It replaced the spreadsheet the kitchen had been working off.
Smart urgency detection: The system distinguishes between standard orders and same-day orders placed late in the day. Late same-day orders bubble to the top of the delivery dashboard automatically.
The Result
The overnight Excel routine is gone. The kitchen starts each morning with a tally pulled from live Shopify data - no manual entry, no copy-paste, no version control issues between the person who made the sheet and the person reading it.
Missed orders went to zero. Drivers are assigned and tracked in the same interface ops is already using. And when something changes mid-day - a cancellation, an urgency escalation - the whole team sees it immediately.
What Made It Work
The technical foundation here is straightforward: Shopify webhooks, a PostgreSQL database, Redis for caching, and Socket.IO for real-time updates. None of that is the interesting part.
What made this project succeed was spending time before any code was written mapping out every exception in their existing workflow. The difference between a Melbourne same-day order and a national delivery. The way add-on products need to be separated from main products in the tally. The edge case where a delivery date isn't explicitly tagged and has to be inferred. The urgency cutoff that changes based on time of day.
Automation built only for the happy path breaks loudly in production. The value is in encoding the real logic - the same logic that lived in someone's head and sometimes didn't make it into the spreadsheet.