Kaplink

For hospitality businesses

What’s eating your week?

First light, before the doors openThe kitchen before serviceOut on deliveries at duskThe oven, late service

Mostly it isn’t the cooking. It’s the rota, the count after close, the supplier orders, the review you haven’t answered. Kaplink uses AI and plain software to take one of those jobs off your plate, on the systems you already run.

Compare notes Try a prompt tonight

Try it · the pass

Which jobs land on you?

These are common admin jobs in hospitality. Put your own hours on the ones that land on you and take away your admin week on a card, with a free prompt to try tonight.

  • 01 · FrontThe Sunday rota
  • 02 · KitchenThe stock count after close
  • 03 · OfficeSupplier orders and price changes
  • 04 · FrontReview replies
  • 05 · OfficeThe weekly numbers
  • 06 · KitchenAllergen and menu updates
  • 07 · OfficePayroll and timesheet checks
  • 08 · OfficeMenu costing

Count my week

Where the time hides

One day, six jobs.

Every service hides admin somewhere. In each case a machine can draft, check or flag. A person still decides.

11:00 The stall sets up, the order still to send
  1. 06:30

    The price that moved

    A supplier changed a price overnight. A machine spots it against last week and flags which dishes it touches.

  2. 09:00

    The Sunday rota

    Holidays, contract hours and who can close. A machine drafts the rota and flags the clashes before it goes up.

  3. 11:00

    From count to order

    Last night’s count becomes today’s order. A machine drafts it from what you use; you check it and send it.

  4. 13:00

    The rush

    Allergen questions and the dish that ran out. A machine can look up your own checked allergen sheet in seconds; your trained team confirms every answer.

  5. 17:30

    The weekly numbers

    Sales, labour and waste from three exports. A machine pulls them into one view you can read in a minute.

  6. 23:00

    The reviews

    The ones you meant to answer. A machine drafts a reply in your voice; you read it and post it.

23:00 The bar at closing, the reviews still waiting

Built in public

We’re proving it on our own kitchen.

Ollie is opening a small pizza kitchen in a garage in Petersfield, with AI helping on the admin behind it: the fit-out, the costings, the orders, the dough plan.

We’re filming it as we go, for a series on YouTube. It is our own business, not a client, and we’ll show what worked and what didn’t.

Where the kitchen stands

How it works

Five steps, one job.

A one-off fix is a fixed price, agreed in writing before anything starts. No retainer.

  1. 01

    Map the job

    How it really runs today, with the people who do it, and a starting measure written down.

  2. 02

    Agree the test

    The check that will show the fix works on your real data.

  3. 03

    Build the smallest fix

    AI and plain software on the systems and data you already have.

  4. 04

    Prove it works

    Your team runs it for real; we run the agreed test together.

  5. 05

    Hand it over

    You own it. One later reading of the same measure shows what changed.

Kaplink has not yet published a client case study. The first will appear after delivery, an agreed impact reading and the client’s written approval.

How it works, in full

Compare notes

Bring one repeated job.

Tell us the job, the systems it crosses and what better would look like. The first conversation is free, and it is about your problem, not a product.

Email Ollie