genfin-demo

GENFIN-AFRICA: Stage-Based Financing Demo System This application demonstrates a cutting-edge model for climate-smart financing, linking smallholder farmer productivity, data-driven scoring, and financial access through the simulation of AI and decentralized ledger technology (blockchain). The core concept is to provide stage-based financing unlocked only when verified farming milestones are met, moving away from traditional collateral-based lending. ✨ Core Features & Value Proposition (In-Scope Demo) The demo system simulates an end-to-end crop financing cycle and highlights key innovations:

• AI Farmer Proficiency Score (FPS) & XAI: An AI-driven score that predicts the farmer's likelihood of successful yield. Users can view the Explainable AI (XAI) factors contributing to the score and risk band.
• Smart-Contract Simulation: The financing lifecycle is managed by a simulated smart contract, providing an immutable audit trail of state transitions (e.g., DRAFT → ACTIVE → STAGE_N_COMPLETED).
• Progressive, Stage-Based Disbursement: Funds are released incrementally across a 7-stage workflow, ensuring they are used productively and only after the preceding milestone is verified.
• Data Governance & Role Restriction: The Insurer dashboard demonstrates restricted views of XAI factors and contract logs to adhere to simulated data privacy policies (POPIA compliance).
• Federated Learning Mock: The architecture includes a simulation of federated learning where clients (farmers) contribute to a model update without transferring raw, sensitive data.
• Mock Integrations: Simulates triggers from external data sources like IoT sensor logs (e.g., soil moisture, pest detection) and file uploads (e.g., soil test reports). 🗺️ Demo Workflow & User Roles The application is driven by different user roles, each with a specific dashboard, to simulate the end-to-end process: | Role | Interface | Primary Function | |---|---|---| | Farmer | Chatbot Mock | Registration, checking STATUS, uploading files, and inputting mock IoT sensor data. | | Field Officer | Dashboard | The trigger point—responsible for validating milestones (e.g., confirming Soil Test Completed or triggering a Pest Event) to unlock the next financing stage. | | Lender/Admin | Dashboard | Portfolio monitoring, viewing the full contract audit trail, and monitoring AI risk scores and XAI factors. | | Insurer | Dashboard | Monitoring policy status and viewing triggers (e.g., low moisture) that could initiate a claim. | | Tester FAQ | Dashboard | Provides context on the solution, detailed testing steps, and a comparison of the Demo Implementation vs. the Actual Solution Goal. | 7-Stage Financing Cycle (Stages 1-7)
• Stage 1: Soil Test (UNLOCKED by default)
• Stage 2: Inputs (Seed/Fertilizer)
• Stage 3: Insurance Premium (Triggers drought policy activation)
• Stage 4: Weeding/Maintenance
• Stage 5: Pest Control (Conditional based on Officer/IoT trigger)
• Stage 6: Packaging
• Stage 7: Transport/Marketing 💻 Technical Stack This demo is implemented as a microservices architecture with separate frontend and backend components. | Component | Technology | Details | |---|---|---| | Frontend | React (JavaScript) | Provides the mobile-first, multi-dashboard user experience (App.js). Uses axios for API communication. | | Backend | Python, Flask | Handles API routing, business logic, and database operations (app.py). | | Database | Flask-SQLAlchemy | Uses a local SQLite file (genfin_demo.db) for persistence and contract state tracking. | | Reporting | Reportlab (Python) | Generates a mock KPI dashboard/report (PDF). | ⚙️ Setup and Installation Prerequisites
• Python 3.8+ (for Backend)
• Node.js & npm (for Frontend)
• Flask and Flask-SQLAlchemy dependencies (listed in app.py)

1. Backend Setup (app.py) The backend is a Flask application.

• Install Dependencies:

  Assuming you use a virtual environment

pip install Flask Flask-SQLAlchemy Flask-Cors reportlab # and any other dependencies

• Initialize the Database: The application requires the database tables to be created first. export FLASK_APP=app.py flask init-db

(This command drops all previous data and recreates a clean SQLite database).

• Run the Backend Server: python app.py

Server will run on http://127.0.0.1:5000 by default.

2. Frontend Setup (App.js) The frontend is a React application.

• Install Dependencies: npm install # Assuming a package.json exists with React, axios, etc.

• Configure API Endpoint (Optional): If the backend is not running on the default http://127.0.0.1:5000, set the REACT_APP_API_URL environment variable.

  Example for Linux/macOS

export REACT_APP_API_URL='http://your-backend-ip:5000'

• Run the Frontend: npm start # Or your equivalent command (e.g., next dev)

⚠️ Important Note on Simulations This is a proof-of-concept demo. The following components are simulated and do not involve real-world interaction:

• Smart Contract: Simulated via database logging and cryptographic hashing in app.py. It does not interact with a real blockchain mainnet.
• AI Scoring: Uses a mock calculation based on static factors (e.g., Land Size, Stages Completed Ratio) rather than a live PyTorch model with real-time data.
• Funds: No real money movement is involved; disbursements are simulated ledger entries.