End-to-end development of machine learning SOTA models (classification and clustering) to classify and segment populations in a large real-world tabular data set (of 10 million individuals with more than 250 million instances). Developed a classification model of Israel's Haredi population which improves recall by 12% and precision by 7% over the best model developed by Israel's central bureau of statisticsCausal analysis of employment, fertility and education using time series and parametric and non-parametric estimationCollaborated with domain experts in social sciences and clients from local authorities and governmental positions.Leading directions for future data science research and management of labelers and analysts.Placed the institute as the leading research institute regarding empirical analysis of the Haredi society in Israel.Generated more than a 1.1 million dollars increase in the company’s revenue in 2023.

End-to-end development of machine learning forecast models for content popularity. Adopted by OpenWeb as the main popularity models in some domains.From translating company’s relevant KPI’s (CTR, retention) to defining evaluation metrics of the project and setting deadlines to successful deployment, passing staging and getting into production environment. Choosing domains for the A/B testing based on statistical analysis and applying analysis and statistical inference on test outcomes for performance monitoring.Using Catboost regression, SentenceTransformers text embeddings and evaluation metrics in recommender systems (NDCG@K). Developed the project in pySpark while utilizing cloud-based technologies through BigQuery and Databricks (cloud environment)

Conducted descriptive and causal inference analysis for Israel's Labor ministry using kernel density estimation, exact matching and multiple types of regression models.Collaborated with domain experts in social sciences and clients in governmental positions.Built infrastructure data for avodata.labor.gov.il and developed a statistical disclosure algorithm for the project

Conducted descriptive and causal inference analysis using linear, Lasso and logistic regression, and time-series models.Developed a predictive structural model to support strategic decision making.Reviewed scientific papers and participated in scientific writing of multiple policy and research documents.