With PulseCart, I built an end-to-end retail demand forecasting system using LSTM and ARIMA models on more than 73,000 records. I improved forecast reliability over baseline ARIMA models and deployed a Streamlit dashboard for forecasts, inventory insights, and scenario analysis.
I also built RetainIQ, a predictive analytics platform that modeled employee attrition, monthly income, and department classifications. I used Scikit-learn models and analyzed an IBM HR dataset with Pandas, NumPy, and Seaborn.
As a Data Science Trainee at Lovely Professional University, I analyzed the California Housing dataset, trained and compared regression and tree-based models, and developed an interactive Power BI dashboard. I’ve also analyzed flight records and presented a Power BI dashboard that contributed to improved on-time arrival rates.
