{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/76acdd8bd4554497a3137a40a6ac8d24\" frameborder=\"0\" width=\"1920\" height=\"1440\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1440,"width":1920,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1440,"thumbnail_width":1920,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/76acdd8bd4554497a3137a40a6ac8d24-41434e9a312e0edf.gif","duration":514.24,"title":"Verdant AI","description":"This Loom demonstrates Vedant AI, an app designed to help Ghanaian farmers diagnose crop diseases and guide actions without needing a constant extension officer. It describes a home screen that pulls a 7-day regional forecast and a warning feed where nearby farmers report pests and diseases from the last 14 days, accessed via phone sign-in without a password. The cocoa and cassava demo shows photo capture that works even with limited signal, and an AI pipeline where three models run in sequence, including a classification model scoring 96% and abstaining when not confident. The app generates farmer-specific weekly recommendations, can filter out chemical names, provides follow-up and history, and includes voice responses for low literacy users, with recording and cost about one cent per scan API usage."}