A collection of smaller, self-contained analytics projects — time series forecasting, Power BI dashboards, and predictive modeling — built alongside my main portfolio project, Quick-Commerce Gap Analysis.
Benchmarked four forecasting models (SARIMA, Holt's Winter, LSTM, Prophet) on 5 years of Ethereum price data, with auto_arima used to select ARIMA parameters. LSTM significantly outperformed the rest (RMSE 135.82 vs. 1031+ for the classical models). Group project.
A Power BI dashboard tracking project health, cost overruns, and employee utilization — recreating internship-style reporting using a Kaggle dataset.
A Power BI dashboard analyzing box office revenue, genre performance, and audience ratings across two decades of movies.
Churn analysis of 10,000 bank customers, originally modeled in Python (Decision Tree, Random Forest, XGBoost, Gradient Boost) and visualized in Tableau.
For my primary, full-depth project — an end-to-end diagnostic analytics pipeline (Python → BigQuery SQL → Power BI) — see Quick-Commerce Gap Analysis.