Hari Hari

I've built data analytics and machine learning projects that turn large datasets into actionable insights, from banking risk analysis to student dropout prediction and streaming-content trends.

In a banking analytics project, I processed datasets of up to 800,000 records and built Power BI dashboards to analyze loan and deposit behavior. My analysis identified 7–10% default rates, higher risk among clients with credit scores below 600, and opportunities to reduce defaults by targeting high-risk clients.

I developed student dropout prediction models using Random Forest and LightGBM, achieving over 95% accuracy across 20,000+ student records. Using attendance, grades, and socioeconomic indicators, I created an early-warning pipeline with F1 scores above 0.85.

I'm a motivated engineering graduate who enjoys Python-based data cleaning, exploratory analysis, visualization, and machine learning. I also analyzed Netflix content data to uncover genre, geography, and content-addition trends using Jupyter Notebook.