Kenya-focused capstone integrating weather, sales, and customer data to predict delivery risks, optimize routing, and provide basic crop advisories.
### **Capstone Project:** Kenya Weather Aware E-Commerce and Agri-Logistics Dashboard.
---
## 1. Background / Problem Statement
A Kenya-based e-commerce and produce-delivery startup operates in multiple cities (Nairobi, Mombasa, Kisumu, Eldoret, Nakuru). The company delivers goods ranging from electronics to fresh farm produce.
The **operations team** has noticed that **weather conditions** especially rainfall and wind significantly impact delivery times, cancellations, and even supply availability. For agricultural produce, weather also affects planting and harvesting windows, which in turn influences stock levels and delivery demand.
Currently, the company lacks a **centralized analytics system** to:
- Monitor and forecast weather conditions for each city.
- Predict operational risks to deliveries.
- Offer simple, data-driven planting/harvest advisories for suppliers.
- Integrate sales, customer, and product data into one interactive dashboard.
>> At the end the student should design and implement an **end-to-end data analytics solution** that automatically collects, processes, and visualizes both **operational data** (orders, deliveries, customers) and **environmental data** (weather forecasts).
The solution should help the company:
- Identify days and locations with **high delivery risk** due to adverse weather.
- Recommend adjustments to **delivery schedules and routing**.
- Provide simple **agricultural advisories** to suppliers based on weather forecasts.
- Track sales, top products, and customer trends alongside weather impact.
---
## 3. Objectives
1. **Data Acquisition**
- Pull product data from a public e-commerce API of your choice.
- Pull customer and order data from a mock API, fakerapi.it/fake-data-download.
- Pull 5-day/3-hour weather forecast data for each target city (Nairobi, Mombasa, Kisumu, Eldoret, Nakuru) from **OpenWeather API**.
2. **Data Processing & Storage.**
- Clean and merge datasets into a relational model (e.g., products, custo …