Zim-First AI Systems Platform for Agriculture

A confidence-weighted view of what Zimbabwe's farmers actually harvested.

This demo runs the platform's real confidence-scoring and yield-estimation logic against the seeded Phase 1/2 pilot — 13,600 farmers across Murehwa, Umguza and Zaka.

Farmers in pilot
Raw self-reported total
Confidence-weighted total

District yield overview

Raw self-reported production versus the confidence-weighted (shrunk) estimate, per district.

DistrictFarmersAvg confidenceRaw total (t)Weighted total (t)

Farmer lookup

See how an individual farmer's self-report is adjusted toward their district's pattern.

Confidence scoring — try it

Enter a self-reported production value and up to two corroborating signals (satellite NDVI proxy, buyer sale record) to see the multi-signal confidence score.

Predict yield — trained model

Loading model\u2026

Upload your own farm data

CSV columns required: farmer_id, district, confidence, est_production_t (nr, yield_t_ha, area_ha optional). Runs the same confidence-weighted estimator, entirely in your browser.