Skip to content
Artifisys
Menu
Start a conversation Talk to us

Agriculture & food supply

Every customer wants its own spec sheet, and a person writes each one

How do we answer customer specification and allergen requests faster when every customer uses its own form?

Book a 20-minute call Free. No demo, no slides.

Figures and rules on this page apply to

Europe

Working somewhere else? The shape of the problem usually travels. The deadlines do not.

What this looks like

A food processor asks for a new blend. Before it ships, their buyer sends a supplier questionnaire, a spec template and an allergen form. A food technologist copies recipe data into each one by hand. The next customer's forms ask the same questions in a different order.

The numbers

Every figure here is someone else’s. Check them.

  • 14allergens that must be clearly shown whenever they are used in food sold in the EU
  • 5,250EU food and feed safety alerts raised in 2024, up 12% on 2023
  • 392EU alert network cases about herbs and spices in 2024, about 4% of all cases
  • €182bnEU food and drink exports outside the EU, each shipment needing documents for its own market

Why it happens

It is not a people problem.

The recipe lives in one system. The customer's questions arrive on the customer's own forms. Nothing joins the two, so a person reads each form and fills it from the recipe. When a second plant joins the group, its recipes and raw materials are recorded in a different way again.

Why your current software has not fixed it

Because it was never built to.

Recipe and ERP software store the right facts. They print their own spec sheet, not the forty customer templates a sales team actually receives. Specification software expects the customer to use the maker's portal. Big food buyers rarely do. So the gap is filled by technologists typing.

Intelligence, plumbed in

A customer's allergen questionnaire, sent as a locked PDF with its own wording, can be read and answered from your recipe data.

We build the reader for your customers' forms, then fill them from records you already keep. Underneath it: a connector, one agreed meaning per field, and a test set scored on your own records.

How we make AI survive real data
  • Connectors
  • A semantic layer
  • Evals you can check

However hard, whatever it is

Seasoning blends are one example. Bring the document your customers ask for most.

We wrote this about ingredient makers because every blend is made for one buyer. The same gap exists for any supplier that fills in customer forms by hand.

  1. 01

    We sit with you

    Days where the work happens, not a workshop in a meeting room. We watch the job get done and write down the shortcuts nobody wrote down.

  2. 02

    We read everything

    Your data, your rules, your vendors and their documentation, and the published research on your sector. We report what is actually in there.

  3. 03

    We break it to first principles

    Not which tool fixes this. What is actually causing it, taken apart until we reach the piece that cannot be divided further.

  4. 04

    Then we build

    Weeks, not quarters. By this point we are not guessing what to build, and guessing is the thing that makes projects long.

What we build

Specific enough to argue with.

Four mechanisms, not four features. Each one is a thing that happens on its own, every day, whether or not anyone remembers to run it.

  • Customer question forms, spec templates and allergen forms read on arrival, in whatever format the customer sent.

  • Each answer drafted from your own recipe and raw-material records, with the source shown for a technologist to check.

  • Raw materials declared differently at two plants flagged, so one ingredient never carries two allergen answers.

  • Recipe changes traced to every customer document they affect, so updates go out before the customer asks.

How you would know it worked

Numbers in your own reporting, not ours.

  • Days from a customer request to a signed-off specification pack.
  • Share of form answers drafted without re-typing recipe data.
  • Number of customer queries about conflicting allergen or ingredient statements.

Straight answers

Where a model is involved, it is scored against your own records first. Accuracy per source, not one flattering average.

The questions this raises

  • Our technologists must sign off every declaration.

    They still do. The layer drafts the answer and shows where each fact came from. The person checks it and signs.

  • We have just added a plant with its own systems.

    Both plants keep their systems. The layer reads each one and shows where the same raw material is described differently.

  • How soon would the queue get shorter?

    Usually within weeks, starting with the customers who send the most forms. More templates are learned as each answer is confirmed.

Every partner sends a different shape

The format belongs to them. The cost lands on you.

Documents arrive from outside in whatever shape the sender chose, and the sender changes it whenever their own system does. No vendor can ship this. The set of formats is specific to who you trade with, which is why it ends up as a script nobody owns.

Bordereaux, carrier files, supplier invoices, delivery notes. Whatever arrives in your inbox in nine shapes is the same job.Send us the messiest one you have.

The spec sheet queue is a reading problem, not a recipe problem.

We build the reader for your customers' forms, then fill them from records you already keep.

See everything we build
  • Software
  • Hardware
  • Ways of working
  • Whole ventures

Is this happening to you? Tell us the size of it.

Twenty minutes. We will tell you honestly whether the numbers justify doing anything about it.