At Nielsen, I analyze audience reach, ratings, and demographics using SQL and Python to explain week-over-week metric shifts to stakeholders. I also investigate viewership anomalies, distinguishing audience behavior from panel changes or upstream data errors.
I led release sign-off by comparing legacy and revised panel processing logic in parallel, quantifying output variance against materiality thresholds. I automated recurring panel reporting in Power BI and built a Python script for monthly content audits that saves 30 minutes per cycle.
On my Real-Time Weather & Air Quality Analytics Platform project, I built an ETL pipeline that ingests live weather and air-quality data from REST APIs into a partitioned AWS S3 data lake. I also created a refreshable Power BI dashboard comparing city-level air quality rankings.
For my Fraud Detection & Risk Analytics project, I built a transaction-level fraud detection pipeline and tuned its classification threshold to favor recall. I also created a Power BI dashboard to surface risk trends and high-risk transactions.
