Open to Work

Analyzing the Signal
in the Noise.

Data Analyst specializing in predictive modeling, business intelligence, and turning complex datasets into actionable strategic narratives.

Projects

rocket_launch $106k netROI (Scenario-based)
show_chart

Customer Churn Prediction arrow_outward

Built a churn prediction system that connects classification models to retention decisions and ROI, focusing on who to contact and how much to spend, not just churn probability.

Recall 96.5%
Precision 53.2%
Random Forest @ 0.20
Python Scikit-Learn Business Analytics Streamlit
trending_up 31.7% improvement
hub

Demand Forecasting arrow_outward

Built a decision-support system that converts multi-horizon demand forecasts into inventory policies, explicitly modeling demand uncertainty using quantile regression

Model XGBoost
WAPE 4.01 %
Horizons 1-4 weeks
Python XGBoost Time Series Inventory Analytics
rocket_launch 80% Revenue
ads_click

Customer Segmentation & Revenue Analysis arrow_outward

Performed end-to-end customer analytics using RFM segmentation to identify high-value customers, revenue concentration, and retention opportunities, delivered through a business-ready Power BI dashboard.

Revenue 80% (High Value)
Retention At-Risk (Rule based)
Tableau / PowerBI SQL Python
schedule -20% Wait
medical_services

Customer Lifecycle Analytics Platform arrow_outward

Built a modular, rule-based decision-support system that transforms raw transaction data into customer insights, risk assessments, and ROI-aware business actions with fully auditable logic.

Decision-Engine Rule Based
Analytics RFM
Python Pandas Analytics Engineering Streamlit
gpp_good 99.9%
lock

Online Payment Fraud Detection arrow_outward

Built a machine learning–based fraud detection system to flag suspicious online payment transactions, with a strong focus on recall optimization under extreme class imbalance.

Precision 88%
Recall (Fraud) 99%
Python LightGBM Machine Learning Streamlit
thumb_up Applied Regression
forum

House Rent Price Prediction arrow_outward

Built a machine learning–based web application to estimate rental price ranges for residential properties across major Indian cities using historical listing data.

Model Random Forest
Rent Analysis City & BHK-wise
Scikit-Learn Python Regression Data Visualization