Two irrigation events, 18 mm total, before 24 July.
Veraison six days early and deficit building. Left alone, this vintage lands riper and heavier than the estate's buyers pay for.
Act before24 Jul
Moves pick to8–14 Sep
KARAZ / Vine
Vine · interpretiveMost estates learn which wine they made a year too late.
A vineyard decision does not move a number on a screen. It moves sugar and acid, which moves style, which moves who will buy it and at what price. The unforgiving part is the delay: by the time the style is obvious in the glass, every point at which you could have changed it has passed.
What actually arrives
Decisions, not dashboards.
Each one names the parcel, the action, the reason, and what it does to the pick date — because in wine the consequence of an irrigation decision is a harvest window and a style, not a moisture reading.
Hold the canopy work. Pressure does not justify the pass.
Modelled disease pressure has not crossed the threshold and the forecast does not carry it there. A calendar pass here spends money and a tractor pass for nothing.
Re-checkIn 96 h
Savedone application
Open the fresher-profile conversation now, not at bottling.
Projected profile for two thirds of the estate sits in a segment your current buyer list under-serves. That is a placement problem best solved eight months early.
Lead time8 months
Share66% of volume
Decision cards illustrate the format and the severity model. Parcels are coded.
Targets per estate
Stated as targets — not as achieved averages.
These are what the vine branch is engineered against and reported on. We do not present them as results across a customer base we do not yet have, and we would treat any supplier who did with suspicion.
| Target | Measured as |
|---|---|
| Water use ≥20% lower | Applied irrigation per hectare against the estate's own prior seasons. |
| Plant protection ≥30% lower | Applications avoided through pressure-based rather than calendar-based timing. |
| Yield held within ±7% | The savings above must not be bought with tonnage. |
| Sell-through +10 pp | Proportion of the vintage placed within the commercial year. |
| Embedded emissions ~10% lower | Per bottle, from reduced inputs and reduced passes. |
Validation runs on a 12-hectare Tuscan estate held as a permanent validation site, with independent verification of the environmental figures through the programme partners described on the evidence page.
The other branch
Cherry runs on the same engine and a different logic.
Wine is interpretive — you read which style the season is producing and match it to a buyer. Cherry is interventionist — you steer six measurable parameters toward a grade, and the arrival date decides what they are worth. Shared sensing and pipeline, separate strategy.
Olive · next
Deliberately not yet
Olive follows the same pattern — its own quality model and market logic on the shared engine. We are not adding it before the two live branches carry their own weight. A platform that claims five crops and understands none of them is the failure mode we are most determined to avoid.