HR Analytics with SQL & Python

This project explores employee attrition patterns using a mix of PostgreSQL for querying and Python for analysis & visualization.
It is designed as a portfolio project — showing end-to-end data handling, from raw CSV → database → SQL analysis → Python stats → machine learning → reproducible deliverables.

---

📂 Project Structure

HR-Analytics-SQL-Python/
├─ queries/
│    ├─ basic_checks.sql
│    ├─ overall_attrition.sql
│ 	 ├─ dept_attrition.sql
│ 	 ├─ salary_analysis.sql
│ 	 ├─ jobrole_attrition.sql
│ 	 ├─ tenure_attrition.sql
│ 	 ├─ overtime_attrition.sql
│ 	 └─ age_gender_attrition.sql
│ 	 
├─ notebook/
│    ├─ hr_analysis.ipnyb
│    ├─ eda_summary.ipnyb
│    └─ figures/
│    		├─univariate/
│    		│ 	├─ age_distribution.png
│    		│ 	├─ attribution_balance.png
│    		│ 	├─ department_counts.png
│    		│ 	├─ monthly_income_distribution.png
│    		│ 	└─ overtime_counts.png
│    		│ 	
│    		├─bivariate/
│    		│ 	├─ attrition_vs_age.png
│   		│ 	├─ attrition_vs_dept.png
│    		│ 	├─ attrition_vs_gender.png
│    		│ 	├─ attrition_vs_jobrole.png
│    		│ 	├─ attrition_vs_maritalstatus.png
│    		│ 	├─ attrition_vs_monthlyincome.png
│    		│ 	├─ attrition_vs_overtime.png
│    		│ 	└─ attrition_vs_tenure.png
│    		│ 	
│    		└──eda_summary/
│    			├─ age_tenure_analysis.png
│    			├─ attrition.png
│    			├─ attrition_gender_maritalstatus.png
│    			├─ dept_and_jobroles.png
│    			├─ monthlyincome_vs_attrition.png
│    			└─ overtime_impact.png
├─ data/
│    └─ ibm_hr.csv
│ 
├─ results/
│    ├─ overall_attrition.csv
│    ├─ dept_attrition.csv
│    ├─ salary_analysis.csv
│    ├─ jobrole_attrition.csv
│    ├─ tenure_attrition.csv
│    ├─ overtime_attrition.csv
│    └─ age_gender_attrition.csv
│
├─ LICENSE	
├─ requirements.txt
└─ README.md # Project Overview (this file)

---

🎯 Objectives

• Understand employee attrition drivers using SQL queries
• Perform exploratory data analysis (EDA) with Python
• Run statistical checks (e.g., overtime vs attrition significance)
• Build a basic predictive model (logistic regression)
• Deliver visuals, CSV outputs, and notebooks in a clean repo

---

🛠️ Tools Used

• Database: PostgreSQL
• Querying: SQL
• Analysis: Python (Pandas, NumPy, Matplotlib, Seaborn, Statsmodels, Scikit-learn)
• Version Control: Git + GitHub

---

📊 Key Findings

• Overall attrition: ~16% of employees
• Departmental differences: R&D had higher attrition vs HR & Sales
• Overtime: Employees working overtime are significantly more likely to leave
• Job roles: Sales reps & lab technicians showed higher attrition
• Income effect: Lower monthly income strongly correlates with attrition
• Predictive model: Logistic regression flagged Overtime, MonthlyIncome, JobRole, and Age as the strongest predictors.

---

📊 Key Findings (Simple Version)

• About 1 in 6 employees left the company during the observed period.
• Working overtime is one of the biggest factors linked to leaving.
• Lower salary employees tend to leave more often than higher salary ones.
• Certain job roles and departments (like Sales and Lab Techs) have higher turnover.
• Younger employees are more likely to quit compared to older ones.

---

📌 Deliverables

• Reproducible SQL queries → queries/
• Query outputs → results/
• Python notebooks with EDA, stats, and model → notebook/
• Figures for visualization → notebook/figures/
• README.md (project overview)

---

🚀 How to Run

1. Clone this repo

   git clone https://github.com/ajx7/HR-Analytics-SQL-Python.git
   cd HR-Analytics-SQL-Python

2. Install requirements

   pip install -r requirements.txt

3. Set up PostgreSQL DB

   • Create a database
   • Import data/ibm_hr.csv

4. Run SQL scripts in queries/

5. Open notebooks in notebook/ for Python analysis

---

📖 Dataset Source

IBM HR Analytics Employee Attrition & Performance dataset (publicly available on Kaggle): IBM HR Analytics Dataset – Kaggle

---

⚖️ License

This project is licensed under the MIT License — free to use with attribution.

---

About

This project was built as part of a portfolio refresh, showing end-to-end data analysis and reproducibility skills.