Days of copying, every single month
The group runs over 100 restaurants across two brands. Every month each restaurant sent its numbers in as Excel files: a P&L, which shows what the restaurant earned and spent, and a scorecard that rates the restaurant and its manager.
Management wanted the same answers every month. Which stores are up, which are down, how one manager compares to another. So someone opened every file, copied the numbers across, built the comparisons and exported the reports. Then the next month came and they did it all again.
What I built
The restaurants keep their Excel files. Nobody had to learn a new tool. I automated everything that happens after the files arrive.
You drag the month’s files in. The app checks them, pulls every number into one database and updates the dashboards: P&L, scorecards, store rankings, month on month, manager against manager. Each manager sees only their own stores. The management pack exports itself, the same way every time.
The AI sits on top of the checked numbers. A manager asks which stores stand out this month or what changed since last quarter, and gets the answer in plain words. The AI explains the numbers. It never does the maths.
“Then the next month came and they did it all again.”
The numbers match finance
Before handover I checked every store’s total, for every month of the quarter, against the finance team’s own figures. Every one matched. Then the app produced the full quarterly management pack on its own.
Every bug I found along the way became a permanent check. An upload isn’t done until the numbers are actually in the live data.