Principal Data Scientist, Predictive Modeling

Responsibilities:

• Design, test and implement recommendation engine models to turn sparse datasets into synthetic datasets.  This includes writing code from scratch.  Must be proficient with reading, manipulating and analyzing big data and writing new code to build models.
• Prove model validity through analytic research (e.g. holdouts) and speak to predictive power internally and externally to clients as needed.
• Use iterative modeling to determine the ideal parameters for the sparse data - e.g. minimum level of completeness, key required datapoints that drive higher predictive accuracy, minimum number of datapoints required for a given predicted variable, etc.
• Using the analytic findings from ideal parameter exploration, consult with internal and external partners on the best way to source the ideal sparse data.
• Support screening and hiring of other data scientists in the mid to long term future.  Support onboarding and development of more junior data scientist staff.  Serve as a subject matter expert for others internally and externally.  Provide technical assistance in predictive modeling methodologies, data manipulation, fusion, and modeling.
• Define and implement a vision for scaling new models across increasingly larger datasets - driving for efficient speed, computer usage, etc.
• Support automation for scaling routinized processes.
• Support pilot programs for R&D purposes.  
• Conduct tactical or strategic analyses to address business and customer opportunities.
• Utilize tools such as Python, R, SPSS, etc. to perform complex data analysis, develop tools for automating procedures.
• Develop, test, and implement high quality, modular python code that can be seamlessly integrated into an existing production system.
• Develop and implement machine learning solutions to leverage big data from internal and external sources.
• Assist with ad hoc analyses and projects.

Qualifications:

• Undergraduate or graduate degree in Mathematics, Statistics, Social Science, Engineering, Computer Science, Economics, Business or fields that employ rigorous data analysis and strong statistical skills.
• 10+ years of relevant data science experience in implementing various types of modeling.  Prior experience with predictive modeling and specifically with recommendation engines is highly preferred.
• Strong skills in Python and relevant packages, including but not limited to pytorch and familiarity with other scripting languages.
• Knowledge of statistical tests and procedures such as Correlation, Regression, Hypothesis Testing, Segmentation Techniques, ANOVA, Chi-squared, Student t-test, and Time Series.
• Familiarity with machine learning and data modeling techniques such as Decision Trees, Random Forests, Incremental Response Modeling, Scoring, SVM, Neural Networks, and Credit Scoring.
• Strong critical thinking and creative problem solving skills.
• Strong planning and organizational skills.
• Strong verbal, presentation, and written skills and English fluency.
• Demonstrated success and effectiveness working in a time-critical production environment.
• Familiarity with SQL, Oracle or other relational database software including manipulation of large data sets.
• Familiarity with BI, Spotfire, Tableau or other data visualization software and techniques
• Proficiency in MS Office suite (Excel, PowerPoint and Word) and/or Google Office Apps (Sheets, Docs, Slides, Gmail).
• Knowledge of Atlassian suite of software including Bitbucket, Confluence, Jira, Hipchat, Crucible and Fisheye.
• Knowledge of Apache Spark ecosystem, Databricks, and AWS