At Reckitt, I cleaned and validated sales data across sources, then used exploratory analysis and K-Means clustering to surface insights on market behavior for sanitizing products. I also fit an ARIMA model to forecast Vanish sales two months ahead.
I designed an interactive Power BI dashboard to share findings and recommendations with non-technical audiences. In SQL Server, I queried the data and cross-checked results against source records for accuracy and completeness.
My Physics background informs my work with statistical analysis and uncertainty. For my UDLAP undergraduate thesis, I built and trained a Mixture Density Network to predict galaxy redshifts while estimating uncertainty for each prediction.
