Explored datasets of Prime Video member browsing preferences, analyzing scroll behavior, time spent by genre, and bounce rates (Apache Spark, AWS Glue).
Collaborated with the data analysis and model training team to refine AI/ML models (scikit-learn, AWS SageMaker), leveraging insights for personalized content recommendations, resulting in 8% increase in watch time during controlled A/B testing.
Team collectively achieved a 15% reduction in data processing time over course of 10 weeks.
