We are solving the biggest problem in the QSR industry

Quick service chains collect far more operational data than they can use. Every location produces a steady stream of signals, yet little of it is acted on. Problems are hard to find, and once found, the fix is even harder to deploy.

Data lives in silos

A restaurant chain runs on a fragmented set of tools: the POS knows about sales, the labour tool about hours, the checklist app about tasks, and none of them talk to each other. Yet most operational problems only become visible across signals: a drop in sales means one thing if staffing was normal, and something else if the opening checks ran late.

Dashboards do not close that gap. They answer the questions someone thought to ask in advance, but fail to address the other problems that arise in daily operations.

Deploying a fix is harder still

Even when a problem is found, the fix still has to reach every affected location: a procedure changed, a checklist updated, a team retrained. Each step lives in a different system, or in none, and nobody tracks whether the change held.

So problems surface late, in the P&L or when guests notice, and the ones that get fixed often come back. Doing this for every location, every day, is the work Kimchi takes on.

How Kimchi works

Kimchi is a team of AI agents. We call them analysts, because that is the job they do.

Every night after close, they read each location's data from that day and flag deviations from its own baseline long before they show up in aggregate reporting. We call this the night shift (the only shift in the restaurant that never needs a roster 😉).

• Breakfast sales −11%

  Lightspeed

• Opening checks done at 7:40

  Checklists

• Reviews: cold food, slow service

  Reviews

Breakfast sales down 11% in 3 locations

Opening runs 40 minutes late, so the first hour of breakfast comes from a kitchen that is not ready.

Area managerHigh urgency

In three locations, breakfast sales are down 11%, which in the POS alone looks like a slow month. But the opening checklist shows those stores finishing at 7:40, forty minutes after the doors open, and reviews from those mornings say the food was cold or slow to come. The late checklist submissions may well get flagged, but as a compliance issue, not as the reason sales are down.

Read together, they are one problem: the morning routine runs late, so breakfast is served from a kitchen that is not ready, and the guests who notice stop coming.

What comes out is not a chart but an opportunity: what is wrong, where, since when and why.

Each opportunity is automatically assigned to the person responsible for it, so every problem has an owner from day one. The previous example lands with the area manager, with the fix already drafted: move the breakfast prep to the evening before, in those three stores.

The analysts find, explain and propose. People decide. We believe AI should make people better at their jobs, and that a human in the loop matters most when a decision changes how a restaurant runs. Less time in spreadsheets means more time for the part of the job that needs people: judgement, leadership and action.

From opportunity to fix

Finding the problem is only half the work. Procedures, training, checklists, audits and stock all run in our own apps, on the same data the analysts read. So the problem and the fix are one workflow.

Take the breakfast problem. Once the area manager approves the fix, Kimchi moves the prep tasks that can be done ahead to the closing checklist of the night before, in all three stores. Not every task can move, but the ones that do can no longer run late in the morning.

Because the fix runs in the same system that found the problem, Kimchi sees whether it held. A fix that works in one location can then become the standard for all of them.

The team

A problem this hard needs deep operational experience, product craft and serious engineering in one team. Kimchi's three founders, equal partners, bring exactly that.

Gaëtan has spent fifteen years in foodservice and multi-site operations, leading strategy and operational transformation at SSP, Avolta, EXKi, Le Crobag, SNCB and Kinepolis. Gaëtan has seen the same problems at chain after chain, and fixed them one chain at a time. At Kimchi, that experience works for the whole industry at once.

Yassine is the glue between product and commercial. He founded his own app studio and later led product at a larger venture studio. At Kimchi, he studies how customers actually work and obsesses over making every workflow as efficient as it can be, so that everyone, from head office to the floor, finds it easy to use and stays connected.

Hussein started writing code at a young age and is the mind behind Kimchi's technology: the platform, the integrations and the AI analysts themselves. With a background in enterprise security, compliance and infrastructure, Hussein keeps all of it secure and compliant, because that is something we never compromise on.

See it on your own data

If you run a quick service chain and want to see what Kimchi finds in your data, book a demo.