AI assistant
Like asking an adviser
who knows your own land.
The assistant sits at ten places in the product and always works on the farm's real data — these parcels, these contracts, this rotation. That is why it can say what will work on field A2 instead of explaining what crop rotation means in general.
Three rules it keeps to.
It decides nothing
A proposal is presented, not applied. Whatever was read out of a contract sits in the form for review — until someone confirms, nothing in the records changes.
It does not do the maths
Areas, soil points and payments come from the system's own calculation, not from the language model. The assistant explains and contextualises — the numbers come from the same source as the table next to it.
It runs on the server
Requests go through the application itself, never straight from the browser. No API key sits in the client, and nothing goes to a service that is not needed.
Where it helps day to day.
Every one of these is built and reachable inside the product. The route is listed with it.
Ask questions about the CAP 2023–2027 and Brandenburg state funding in plain language. The assistant answers with your own farm in view rather than reciting a leaflet.
Next to every figure and every technical term an explanation is available: what it is, how it was calculated and what it rests on.
Upload a lease or plough exchange document as PDF or photo. The assistant extracts contract type, parties, term, rent and the parcels named in it and presents them for review — nothing is applied without being asked.
Reconcile names from a contract against the existing list of people. Spelling variants, salutations and abbreviations are recognised instead of creating a second record for the same person.
Derive a proposal for the working day from work steps, machinery, capacity and the weather.
Generate a planting plan proposal that considers rotation, parcel layout and timing together.
Put two plan variants side by side and explain the difference in words, not only in numbers.
Generate the customary work steps for a crop as a template, which is then adapted to the farm.
Recognise species and sex on camera images and summarise the captures of a period — with thousands of triggers a week, that is the difference between an evaluation and an archive.
Turn a scribbled sentence into a task with a title, a deadline and a reference — and at the end of the season, an evaluation of what was left undone.
Ask first, sign up later.
The advisory chat is there from the first day of the trial — with your own parcels in it.