📊 Data Analytics & Machine Learning Portfolio

Hi, I'm Harsvardhan Rajgarhia 👋
3rd year B.Tech CSBS student | Aspiring Data Analyst & ML Engineer

This repository contains my end-to-end Data Analytics, Business Intelligence, and Machine Learning projects using Python, SQL, Power BI, Scikit-learn, Streamlit, Excel, and more.

---

⚡ Tech Stack

Programming & Analytics

Business Intelligence & Visualization

Data Engineering & Storage

---

❓ About Repository

This repository showcases my analytics and ML workflows, covering:

• 🔄 Data Cleaning & Preprocessing
• 📊 Exploratory Data Analysis (EDA)
• 📈 Interactive Dashboards (Power BI / Streamlit)
• 🤖 Machine Learning Models (Classification, Prediction & Insights)
• 💡 Business Recommendations
• 📘 Documentation + Setup Guides for Each Project

Each project folder contains:

• 📂 Raw & cleaned datasets
• 📘 Documentation PDF
• 📈 Dashboard files
• 🧪 Jupyter notebooks
• 🧹 requirements.txt
• 📊 Screenshots

---

🚀 Projects

---

1. 📡 Customer Churn Analysis (Power BI + Python + SQL + ML)

End-to-end analytics & machine learning project

• Tools: Python (Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn), SQL (DuckDB), Power BI
• Dataset: Telco Customer Dataset (~7,043 records)

🔍 Key Insights

• Overall churn: 26.5%
• Month-to-month contracts → churn ~42%
• Fiber Optic users → churn ~41%
• Electronic Check → highest churn among payment methods
• New customers (<6 months) → highest churn risk

🤖 Machine Learning Model

• Logistic Regression & Random Forest
• Accuracy: ~83%
• ROC-AUC: ~0.87
• Exported churn probabilities + risk buckets for Power BI integration

---

2. 👥 Employee Attrition Analysis (Python & Streamlit)

• Tools: Python (Pandas, Matplotlib, Seaborn, Plotly), Streamlit
• Dataset: IBM HR Analytics (~1,470 employees, 35+ features)

🔍 Key Insights

• Overall attrition: ~16%
• Sales department → highest attrition (~20%)
• Overtime employees → 3x more likely to leave
• Low-income group (2.5k–5.5k/month) → higher attrition
• Younger age group (25–35) → higher attrition

---

3. 💹 Sales & Profit Analysis Dashboard (Power BI)

• Tools: Power BI, DAX, Power Query
• Dataset: Simulated 20K+ sales transactions

🔍 Key Insights

• Profit Margin ~29%
• South region → top revenue (~₹700K)
• Product P010 → best performer
• Bangalore Hub → highest sales

---

🌟 Future Work

• Predictive Sales Forecasting
• Time-Series Analysis
• Deep Learning Projects
• Portfolio Website

---

👤 Connect With Me

📧 Email: [email protected]
💼 LinkedIn: **https://www.linkedin.com/in/harshvardhan-rajgarhia-ba6