Piyush Bagade

I’m a Research Intern focused on turning complex biomedical and business data into reliable, actionable results. At NIELIT, I analyzed 100,000+ ECG heartbeat records across 48 patient datasets through cleaning, filtering, normalization, and quality validation.

I run exploratory data analysis to uncover ECG pattern insights and class distribution issues. I also apply resampling techniques to address imbalances, then build automated Python pipelines for heartbeat segmentation, feature extraction, and class classification—reducing manual effort and improving workflow efficiency.

I evaluate models using precision, recall, F1-score, confusion matrices, and training-validation performance metrics to ensure dependable outcomes. I’ve also validated and deployed trained ECG classification models on NVIDIA Jetson Nano edge hardware to enable real-time inference for portable healthcare monitoring.

Beyond ECG work, I enjoy analytics that drive decisions. I built SQL-based analysis for the Chinook Music Store (customer behavior, churn/retention, and revenue using joins, CTEs, subqueries, and window functions), and I created interactive Power BI dashboards in areas like music analytics and call-center performance with KPI measures, reporting, and drill-through visualization.