sbouazzi/ardhiq

AI-powered farmland analysis for Tunisia # ArdhIQ (Spec-based MVP) This repo contains a **Next.js web app**, a **Flutter mobile app**, and a **Python backend** implementing the ArdhIQ spec you shared: - Conservative, skeptical land-offer analysis for Tunisia - Fixed output template + 0–100 scoring breakdown ## Repo layout - `backend/` – FastAPI service (SQLite + SQLAlchemy + OpenAI + optional OCR) - `web/` – Next.js MVP web app (landing, analyze, result, history, admin) - `mobile/` – Flutter app (API client + UI to paste offers and view reports) - `docs/` – spec documents copied from chat ## Quick start (backend) 1. Create `backend/.env`: ```bash OPENAI_API_KEY=sk-... MODEL=openai/gpt-5.2 ``` 2. Run: ```bash cd backend python -m venv .venv && source .venv/bin/activate pip install -r requirements.txt uvicorn app.main:app --reload --port 8000 ``` 3. Test: ```bash curl -X POST localhost \ -H 'Content-Type: application/json' \ -d '{"offer_text":"أرض فلاحية للبيع 7 هكتارات 1200 شجرة زيتون بئر 45 متر 470 ألف دينار"}' ``` ## Quick start (Web) 1. Configure `web/.env.local` from `web/.env.example`. 2. Run: ```bash cd web npm install npm run dev ``` 3. Open . ## Quick start (Flutter) 1. Install Flutter (local machine). 2. From `mobile/`: ```bash flutter pub get flutter run ``` 3. Set backend base URL inside the app (Settings). ## Notes - The authoritative analysis rules live in `docs/ardhiq_system_prompt.md`. - The backend **instructs the model to output exactly** using that template.