• Improved the ranking algorithm at Amazon Gifting, a curated catalog of products designed to inspire gift ideas, by using Java and Spring
• Delivered code in a microservices-based model that dealt with data of millions of products, using the big data library Apache Spark and ElasticSearch, to rank them in their correct categories in AWS S3
• Aided in the ranking of products by implementing Thompson Sampling, an exploration/exploitation probability model, which brought less-popular products up in the rank
• Collaborated with a team of 9 on ideas for new ways of optimizing product ranking and curating
• Contributed to Correction of Errors during the peak, Q4, Thanksgiving period
• Led A/B-testing of the experiment which produced 3% increase in sales within 1 week after the end of the project, in key markets such as UK, Germany and France