End-to-End Insurance Risk Analytics & Pricing Engine for AlphaCare Insurance Solutions (ACIS) in South Africa, featuring EDA, Data Version Control (DVC), A/B Hypothesis Testing, and predictive modeling (XGBoost/Random Forest).
# AlphaCare Insurance Risk Analytics
An end-to-end data science and analytics pipeline designed to analyze historical insurance claim data, optimize marketing strategy, and identify low-risk customer segments for AlphaCare Insurance Solutions.
## 📊 Project Overview
This repository contains the exploratory data analysis (EDA) framework and data engineering practices built for Task 1 and Task 2 of the interim milestone. The core objective is to calculate underwriting KPIs (Loss Ratios, Profit Margins) across geographical and demographic categories to spot optimization opportunities.
## 📂 Project Structure
```text
insurance-risk-analytics/
│
├── data/
│ ├── insurance_data.csv # Raw historical dataset (git-ignored)
│ └── insurance_data.csv.dvc # Data Version Control tracker file
│
├── notebooks/
│ └── 01_eda.ipynb # Core analysis, calculations, and plotting panels
│
├── src/
│ ├── __init__.py # Makes src importable as a module
│ ├── data_loader.py # Safe, robust custom dataset loading pipeline
│ └── eda_utils.py # Reusable styling and custom visualization functions
│
├── .gitignore # Production-ready file exclusion configuration
└── README.md # Project documentation (You are here!)