Alexandra Darmon

Experience

1. Research Scientist - R & d

   Meta

   Apr 2022 — Present

   Ranking Models: Working on several projects to improve the efficiency of the ads ranking models. Examples include the assessment and reduction of the cold start problem for new campaigns using smart embedding, the optimization of the initial stages of ads retrieval leveraging Adaptive Experimentation and Bayesian Optimization, improving the mechanism to handle Nulls in input features by correctly specifying the missing data mechanisms. Causal Inference: Leading development of Bayesian hierarchical modeling for the internal causal ad delivery experimentation platform. A cross functional team with engineers, data- and research scientists. The framework is used by multiple teams and resulting in significant variance reduction with higher degree of calibration of credible intervals. Designed metrics for teams to evaluate the long term causal impact of our ads and commerce experience on user behavior. Causal Modeling: Developing statistical methodology for assessment of long term viability of causal ad delivery models. • AutoML: Optimizing the lifecycle of ML models in Integrity and Financial Integrity.

Skills

• Spanish
• Financial Modeling
• Entrepreneurship
• English
• Financial Analysis
• French
• Photoshop
• Microsoft Office
• Options
• Marketing
• Economics
• Microsoft Excel
• Market Research
• Community Management
• Business Strategy