Tasting sessions ran at one location while we built it — chefs tried the prep lists the system produced and told us what felt wrong.
Case 02 · Catering
Prep planning ate a full day a week at every location. Now it mostly runs itself.
A regional catering and events company running more than a dozen locations.
Outcome · most of the prep time, gone- Forecasting
- Supplier orders
- Allergen checks
- Every location
The issue
Where it hurt.
Prep planning was a weekly ritual: head chefs cross-checking the event book, walking the cold store, and scribbling supplier orders on the back of a menu. It took most of a day a week at every location, and it still missed — a fair bit of food thrown out, a few inventory mistakes a week, and a long lag between realising they needed something and the supplier actually getting the order. Their operations lead was honest with us: she didn't think this could be automated. She'd spent the better part of a year looking at off-the-shelf tools, and none of them could handle the way her chefs actually plan. We came in right at the point where she was ready to build something that fit, rather than license something that didn't.
Discovery
What five days of looking found.
- 01
We started with a few months of booking data, lined up against actual sales and what was received at the door — a handful of menu types covered the vast majority of bookings.
- 02
A week on-site with the head chefs came next: walking the receiving dock, the cold store, the prep stations, writing down every rule that only lived in someone's head — supplier minimums, allergen swaps, what's in season.
- 03
From there it was a matter of mapping every ingredient, every supplier, the lead times, and the point at which something starts going to waste.
How we worked
Side by side, not over tickets.
-
01 -
02 There was a standing early-morning call with their operations lead through the first few months — the one time of week everyone was on the same line.
-
03 Site visits rotated through every location during rollout so every head chef met the team face to face.
The build
What runs now.
- 01
The system looks at what's booked, the usual walk-in numbers for that day, and the type of event, then produces a prep list per shift, broken down by station, with the portions worked out around each chef's allergen and dietary swaps.
- 02
Supplier orders get drafted every morning with waste limits built in — if the pattern says a chunk of something usually gets binned, the order quietly comes down by that much. Allergen swaps get checked automatically.
- 03
Prep lists land on the kitchen tablets first thing, and supplier orders go out shortly after — including the few suppliers who still only take a fax.
- 04
Every morning the owners get yesterday's numbers per location waiting for them.
Outcome
What it's worth.
most of the prep time, gone- The chefs got roughly a day a week back each — most of it went into actually developing menus.
- Nobody in the head-chef ranks has left since it launched, which wasn't the case before.
- Saving
- Roughly half a million a year recovered once you add it up — mostly far less food in the bin and far fewer inventory mistakes.
- Speed
- Getting an order to a supplier went from about a day and a half to under an hour. Prep planning went from most of a day a week at each location to a couple of hours.
- Reliability
- Inventory mistakes went from a few a week to almost none. The allergen check has already caught swaps that would otherwise have gone out the door.
Where it is now
It runs on its own day to day now. We're still involved — menu changes fold in every couple of weeks — and their operations lead brings new locations online herself off a guide we wrote together; the last couple went live without us touching the build.
Your next step
Your version of this story.
Thirty-minute call. We'll find three flows worth automating before we hang up. No deck. No "discovery proposal." Just the work.