Applied Machine Learning

Created 9 years ago in Technology / Other

At Amazon, I was responsible for automating model building and scoring of 3k+ propensity models and 1k+ for downstream impact analysis at regular cadence.

Some of the machine learning models we automated were propensity of purchase by a customer in product category, sub category, by digital marketing channel, by event such as propensity to purchase on Black Friday/Halloween/Cyber Monday etc.

Also we build models predicting downstream value of an event such as sign up for Amazon Credit Card or First purchase in a category etc

At Microsoft Bing, I build Machine Learning models such as the first version of binary classifier for automatically classifying web queries as commericial/non commercial.

I also worked on multiclass classifier for classifying web queries to Bing Shopping product taxonomy and developed algorithms for metadata(brand, model, spec attribute etc) identification within retail queries.