OrchardHQPhotograph the symptom; get a probable cause and the right follow-up.
Not a generic plant-disease app. A vision model tuned with the tree's variety, the disease vocabulary for that fruit, and an Irish climate prior — so canker, scab and silver leaf rank above fireblight.
1. The flow on the phone, four taps end-to-end
Standing at the tree, open it on the phone. Take a tight shot of the symptom — leaf, fruit, or bark — and let the diagnoser do the rest. The result lands as an observation on the tree's timeline, with a tickable list of follow-up tasks already drafted for you to accept.
Prefer to pick? Use the disease picker
If you already know what you're looking at, skip the AI: tap + Observation, choose disease or pest, pick from the fruit-scoped list (apple scab, powdery mildew, canker, brown rot, woolly aphid, codling moth…), set severity 1–5. Attach the photo manually.
2. How the model gets it right for an Irish orchard
We send the photo to the vision model along with the tree's variety, the diseases known to affect that fruit, and a local climate prior so canker, scab and silver leaf rank above fireblight in Ireland. Severe or unusual results carry an explicit "consult an expert" caveat. The observation appears in Recent with a diagnosis tag, and the follow-up suggestions are tickable on the same screen.
3. Back at the desk: history and pressure
On the web app, the tree's Activity tab shows the diagnosis chain: this year's photo + diagnosis → this year's clean-up action → next spring's follow-up inspection task. The links accumulate into per-tree evidence on whether the diagnosis was right and the intervention worked.
place /guide/images/diagnose-desk-chain.jpg · 1200×680 · per-tree Activity tab on the web app showing a multi-year chain: diagnosis → cleanup action → follow-up inspection → next diagnosis, linked visually
Pressure overlay across the orchard
Switch to the orchard-wide view: the disease pressure overlay maps incidence spatially for the selected disease and year range, so you can see whether scab is concentrated in the damp corner or spreading. Sits next to the yield heatmap and the pollination overlay in the analytics tab.
Where it's strongest, and where it isn't yet
Diagnose is best on common mid-stage presentations of the diseases and pests in the vocabulary. Rare conditions, very early symptoms, and anything visually subtle are areas we're still working on. A few capabilities — annotated overlays, fruit identification, broader damage-cause coverage — are on the way but not in your hands yet.
Photos go to a third-party AI provider; that's disclosed in our terms. Diagnose has a generous trial allowance on the free tier and sits under the Smallholder plan for sustained use.
The diagnoser's job isn't to replace your judgement — it's to correct your first instinct when the obvious-sounding answer is wrong. The walnut-leaf post is the clearest worked example: walnut leaf, late October →
From the orchard
A guide explains; a walkthrough shows. Posts from Frank's own orchard in Gort:
Don't see what you need? hello@orchardhq.app — we'd like to hear about it.
Try the diagnose flow on the demo orchard.
Try the demo → Request access →