Skip to main content
Start a projectPlant Health Clinic interfaceExplore the workPlant Health ClinicView case study
All work

Applied AI2026

Plant Health Clinic

Farmers send photos, symptoms, and voice notes from the field. Vizier drafts a diagnosis. A SKUAST-K expert reviews it and sends the advice.

  • Product design
  • Farmer app
  • Expert workflow
Plant Health Clinic expert case review in a synthetic demonstration.
Plant Health Clinic farmer advisory in a synthetic demonstration.
Synthetic demonstration · 2 October 2026Art-directed stills · 18 October 2026AI draft responses are simulated.Explore screens About this material

Care, a little closer.

Farmers use the Android app to send a plant photo, voice note, and short symptom form. Plant pathologists at SKUAST-Kashmir review each case in a web dashboard, confirm the diagnosis, and send a treatment plan.

The Vice-Chancellor launched the system at the Directorate of Extension. It runs on Cloudflare, so the institution can handle cases without maintaining a server.

Role
Product + full-stack + AI
Interfaces
Farmer app and expert dashboard
Institution
SKUAST-Kashmir
Public milestone
App inaugurated · 18 May 2026

How it works.

Vizier uses SKUAST-K pathology material to prepare a draft.

  1. SKUAST-K plant-pathology material is stored in one searchable library.

  2. The system retrieves passages that match each case.

  3. Vizier uses those passages to prepare the draft diagnosis.

  4. Treatment suggestions are limited to approved or locally available chemicals.

Vizier is the AI component in this workflow. It retrieves relevant reference material and prepares a draft; SKUAST-K experts decide what advice is sent.

Inside the system.

The flows, sources and review points behind the interface.

The diagnostic loop

A field report becomes a reviewed advisory. The advice returns to the farmer; expert corrections return to the reference library.

  1. FarmerPhotos, audio and symptoms
  2. Case created
  3. VizierDraft diagnosis
  4. ExpertTriage and confirmation
  5. Official advisoryTreatment plan sent to the farmer

Evidence used for a draft

The AI draft is prepared from the case and retrieved pathology material. The expert reviews the diagnosis and treatment plan before advice is sent.

  1. CaseImage and symptoms
  2. Retrieve evidenceSimilar cases and SKUAST advisories
  3. Reference librarySKUAST pathology in Vectorize
  4. Reason over evidenceMultimodal Vertex AI
  5. Draft diagnosisRegion-approved chemicals only

Built for the case and the expert.

Let the farmer show the problem.

Photos, voice notes and symptom details form one case. Photography guidance appears in the farmer app, while offline support and notifications are part of the Android workflow.

Keep the expert responsible for the advisory.

Vizier prepares a draft from the case and relevant SKUAST pathology material. A plant pathologist reviews the evidence, confirms the diagnosis and sends the treatment plan.

Keep corrections useful.

Expert corrections are retained for testing and connected to the reference loop. The review record remains part of the system, alongside the original field evidence.

Built around
the job.

Crop diagnosis system.

Technology

React 18 · TypeScript · Capacitor · Cloudflare Workers (Hono) · D1 · R2 · Vectorize · Vertex AI · Firebase Auth + FCM

In the news.

Coverage related to Plant Health Clinic.

Rising Kashmir

“farmers can directly interact with SKUAST-K experts”

Coverage of the e-Plant Health Clinic app’s inauguration at SKUAST-Kashmir.

Read story

From the product.

Actual farmer-app captures from the Plant Health Clinic review build, 20 September 2026. Medical and agricultural advice is not provided through this portfolio viewer.

Next project

Another product, workflow and set of decisions.

SKIIE public ecosystem page with programme links and published outcome figures.
SKIIE

The site lists SKIIE startups, programmes, events and facilities. Staff manage the content through an admin panel.

Have a workflow or product in mind?

Start a project