Boston, Massachusetts, United States
• Managed the product recommendations team to develop personalized candidate generation and scoring models
• Built a transformer-based large scale product recommendation model using state-of-the-art NLP techniques.
• Researched, developed, AB tested, and deployed to product the Time Informed Calibration Python package, which uses time series models to improve the predictions of probabilistic and regression models.
• Developed software, applications, and AB tests for the Share of Voice platform, which uses optimization algorithms to balance marketing spends between established and emerging product categories.
• Led the integration of workflows between the SWE and DS teams to produce a model auto-retraining platform.
• Collaborated with engineering and analytics teams to develop an application that identifies potential new customers using natural language processing, XGB, and neural network models
• Supported team product discovery and software development efforts leveraging technologies including Google Cloud (including Google Bigquery), Hive, Spark, Airflow, Mamba