jenniehowel7/Agriochestrator-Kenya

AI Autonomous Hackathon # AgriOrchestrator Kenya AgriOrchestrator Kenya is a premium, hackathon-ready Flask web platform for Kenyan smallholder farms. It combines a hierarchical multi-agent engine, OpenAI-powered advisory intelligence, autonomous simulation, and a model lab where you can hot-swap a trained model with a visible model card. ## Why this can win - Solves a real Kenya problem: low-resilience farming decisions under weather and pest pressure. - AI-rich design: multi-agent planning + LLM copilot + predictive model layer. - Demo-friendly: premium landing page, authentication, role-aware user pages, autonomous operations, and decision trace explanations. ## Core Features - Multi-agent autonomous loop (Perception, Forecast, Reasoning, Action, Optimizer). - Farm telemetry simulation for Kenya county profiles. - AI Farm Copilot using your OpenAI API key (live or offline fallback). - Model Lab with model-card powered details and easy Colab replacement flow. - Beautiful premium dashboard with charts and operational alerts. - M-Pesa transaction simulation tab with persistent ledger and payment outcomes. - Real public dataset pipeline for Kenya weather and agricultural indicators. - Role-based login with farmer profile management backed by SQLite. - Offline-first queue for poor-connectivity operations with manual sync flush. - FastAPI mobile backend for authentication, farmer profiles, and sync operations. - Universal Flask frontend with responsive landing, auth, dashboard, profiles, payments, data hub, and sync pages. - Real model lineage visibility (artifact path, hash, model version, trained years, feature list). ## Recent Updates (March 2026) - Added role-aware access control across all major pages (farmer, officer, admin). - Added offline-first sync queue workflows for profile and payment operations. - Added FastAPI mobile backend for auth, farmer management, and sync endpoints. - Expanded real-data utilities with dataset verification support via `scripts/verify_data.py`. - Im …