Transforming raw data into actionable business intelligence
Built an interactive Tableau dashboard to visualize P&L performance, rolling profitability trends, and scenario modeling for executive decision-making. The tool replaced static Excel reports, reducing monthly close analysis time by 30%.
Built an ETL pipeline that automated daily sales reports, KPI tracking, and underperformance alerts. Reduced manual work by 85% with scheduled workflows.
Analyzed 10M+ user events to identify high-conversion purchase paths. Design a scalable PySpark pipeline on Databricks (Delta Lake) to process real-time behavioral data, improving analytics latency by 35%.
Deep Learning model to predict whether charitable donation applications will be successful. Using historical application data, we frame this as a binary classification problem and benchmark our model against a 75% accuracy target.
SparkSQL analysis of home sales data to determine key metrics. Created temporary views, partitioned data, and optimized queries through caching. Compared performance between cached, uncached, and partitioned large datasets.
Machine learning model to assess loan applicant risk using supervised learning techniques with imbalanced classes.