About AgroVision AI

The challenge

Crop diseases reduce yield and quality and raise costs. When a farmer notices spots or yellowing on a leaf, timely expert advice is not always close at hand, and many problems — diseases, pests, nutrient shortages and weather stress — look alike. A wrong guess can mean an unsuitable treatment, wasted money or a delay in getting real help.

Where computer vision helps

A smartphone photo is an easy way to record what you see. A computer-vision model trained on labelled leaf images can recognise visual patterns and suggest which known condition a photo most resembles. It cannot see everything an expert sees, so we pair it with honest confidence information, an “uncertain” result when it is unsure, and guidance that agricultural reviewers maintain.

Our mission

Apply computer vision, machine learning and verified agricultural knowledge to help farmers recognise potential crop-health problems earlier and make better-informed decisions — including knowing when to call an expert.

Nigeria first

Our first market is Nigeria. Public reference datasets do not reflect Nigerian fields, varieties, lighting or phone cameras, so a core part of our work is collecting consented local images with expert-confirmed labels and testing the model on them before making any accuracy claims.

What we do not claim

• AgroVision AI does not replace agronomists, plant pathologists or extension officers.
• It only screens tomato leaves today; other crops are not supported yet.
• We do not publish accuracy figures until the model has been tested on local field images.
• We never give pesticide doses, mixing instructions or product schedules.