Bipin Dahat

I've built analytics and machine learning projects spanning e-commerce sales, used-car price prediction, and an offline RAG-based AI teaching assistant. My work turns messy datasets into clear insights, predictive models, and interactive dashboards.

For a Flipkart e-commerce sales dashboard, I cleaned and analyzed 20,000+ product listings across 25+ categories using Python and pandas. I built a Power BI dashboard with KPI cards, category slicers, and six visuals, using price and discount features to uncover pricing insights.

At InternPe, I analyzed structured datasets and developed supervised machine learning models, handling preprocessing, feature engineering, validation, and evaluation through ROC-AUC, precision, recall, and confusion matrices. At upSkill Campus, I automated data manipulation and preprocessing tasks with Python.

I've also reduced car-price prediction RMSE by 23% versus a baseline linear regression model and developed a fully offline RAG teaching assistant that creates timestamped, citation-grounded Q&A from lecture videos.