joetal24/Smart-Shamba

# 🌾 Smart-Shamba: Local Analytics Lakehouse for Agricultural Decisions A complete, local-first analytics lakehouse that provides data-driven crop planting recommendations for Ugandan smallholder farmers. --- ## 📖 Project Overview **Problem:** Farmers make planting decisions based on intuition, leading to losses from weather volatility and market fluctuations. **Solution:** Smart-Shamba integrates weather, vegetation health (NDVI), and market price data to generate explainable, district-level crop recommendations. **Tech Stack:** 100% free-tier tools running locally - **Orchestration:** Mage.ai - **Storage:** DuckDB - **Transformation:** dbt Core - **Visualization:** Streamlit - **Language:** Python & SQL --- ## 🏗️ Architecture ``` External APIs → Mage.ai → DuckDB (Raw) → dbt (Transform) → Streamlit ↓ Business Logic ↓ Crop Recommendations ``` ### Data Flow (ELT Pattern) 1. **Extract & Load:** Mage pipelines pull data from APIs and load into DuckDB raw tables 2. **Transform:** dbt models clean data (staging) and apply business rules (marts) 3. **Analyze:** Streamlit dashboard visualizes recommendations --- ## 📂 Project Structure ``` smart-shamba-lakehouse/ │ ├── requirements.txt # Python dependencies ├── setup_project.py # Automated setup script ├── .env.template # Environment variables template │ ├── mage_load_weather.py # Weather data loader ├── mage_load_prices.py # Price data loader ├── mage_load_vegetation.py # Vegetation data loader │ ├── warehouse/ │ └── duckdb/ │ └── agri_analytics.db # DuckDB database file │ ├── dbt/ │ ├── dbt_project.yml # dbt configuration │ ├── profiles.yml # Database connection config │ ├── dbt_sources.yml # Raw data sources │ │ │ └── models/ │ ├── staging/ │ │ ├── dbt_stg_weather.sql # Clean weather data │ │ ├── dbt_stg_prices.sql # Clean price data │ │ └── dbt_stg_vegetat …