Senior Data Scientist
CurrentProvide data analysis, statistical insights, and predictive modeling expertise to support the A Place for Mom Sales organization. Deliver key insights, automation, visualizations, and machine learning models in order to empower the sales organization to assist families we work with and to drive conversion of leads to move-ins.• Developed the Move in Probability Score model (MIPS), utilizing a PySpark Random Forest Classification algorithm to predict the probability of lead conversion from initial connection to a placement. Conducted extensive data exploration, feature engineering, and optimization to identify key variables and insights to enhance model performance. Model utilized by 350+ sales employees to prioritize quality of leads in combination with Smart Follow Up List.• Lead statistical analysis and assist in code development of Smart Follow Up List. This tool uses the MIPS model probability score and tree-based decision-making to create a daily list of highest quality leads for each sales advisor. This is predicted to create a net $15-25 million impact after one year.• Lead 5 AB test pilots planning, reporting, and statistical analysis. Successful pilots in combination resulted in a transformational impact to overall connection rate of 1.4 percentage points leading to a $1.68 million increase in yearly revenue and a $375,000 a year in vendor cost reduction.• Create and maintain automated workforce management scheduling and adherence tool utilizing Python, Fivetran magic folder, and Tableau. Process allows for automatic ingestion, transformation, cleaning, storage and visualization of manually created schedule data in combination with database status data.• Integrated email and Google Drive into Databricks using Python, Google API, and Amazon Simple Notification Service (AWS SNS) allowing for reporting automation that saves 80+ work hours per week.