Eyasdm/agrosense-ios

Offline-first iOS crop advisory app for farmers — SwiftUI, SwiftData, inspired by IndabaX Sudan ML Hackathon (4th/132) # AgroSense Offline-first iOS crop advisory app for smallholder farmers. > 🌱 **Status: Early exploration** — core models and views built, on hold while prioritizing BizCore Mobile. --- ## The Problem Farmers in Sudan and rural Indonesia face unpredictable dry spells with no accessible advisory tools. Most existing solutions require internet — which rural farmers don't have reliably. --- ## Origin This app is the direct continuation of work from the **IndabaX Sudan ML Hackathon** (December 2025), where a dry spell prediction model placed **4th out of 132 teams**. AgroSense explores the question that competition raised: how do you get that prediction to a farmer in the field? → IndabaX model repo --- ## What It Does - 🌦️ Dry spell risk dashboard based on current season — no internet needed - 📚 Offline crop library with planting calendars and full growing advisories - 🔖 Bookmark crops for quick offline access - 📷 CoreML crop disease identifier — planned (on-device, no internet required) --- ## Why Offline-First The target user has unreliable internet. Every feature that requires a server call is a feature that fails in the field. AgroSense works from the moment it is installed — no account, no login, no connection required. --- ## Tech Stack `SwiftUI` · `SwiftData` · `MVVM` · `iOS 17+` No backend. All crop advisory data is bundled in the app and seeded into SwiftData on first launch. --- ## Current State Core data models, seed data, and initial views are built. The project is paused to focus on BizCore Mobile — a client-facing iOS app with a more immediate real-world need. AgroSense remains a planned next step for bringing the hackathon research to end users. --- ## Author Eyas Mohammed · eyas.dev · github.com