Yang Song, Ph.D.

Data Engineering & Platforms

As a Software Engineer at Groupon from February 2016 to May 2018 in Seattle and Dublin, I focused on backend development and optimization of cloud-based data pipelines that directly supported an in-house ML recommendation system, enabling personalized customer filtering, offer and incentive selection, and delivery of daily marketing campaigns to over 6 million users. I developed core RESTful APIs using Java 8 and the Play framework to handle push notifications, while leading the migration and redesign of batch processing jobs from Hive to Spark 2.4 on Hadoop clusters, achieving 10x performance improvements via direct YARN submissions and deprecating AWS SWF/SQS workflows, which honed my skills in cloud migration, resource optimization. Owning the architecture of data pipelines across Hadoop, Cassandra, and Redis, I reduced runtime by over 50%, storage by more than 80%, and accelerated data writes by 2x without impacting SLAs, collaborating closely with data engineering teams to ensure seamless integration with ML models for recommendation engines. This hands-on expertise in Scala, Maven, Hive, Spark, and AWS, combined with e-commerce domain knowledge, equipped me to architect cost-effective, resilient cloud solutions, provide technical guidance on big data and ML integrations, and support enterprise migrations.