Statistical Modeler

Job Duties

Own multiple data science projects to deliver end-to-end solutions, including the development of KPIs, new insights through the combination of datasets, and maintenance of ongoing analytics datasets. Collaborate closely with the data science, product, and engineering teams to improve existing product and build new capabilities including automating data-driven tasks, developing algorithms and features to guide optimization and expansion opportunities for the business.

Work cross-functionally within the organization to deliver high-impact, integrated business solutions.

Linear Advertising currently accounts for a large portion of Tatari's total revenue. As a subject matter expert on the Data Science team for linear clearance campaign management, create and improve the Linear Clearance Predictor that allows Tatari buyers to quickly project how much budget each client will be able to spend. Use predictive modeling, SQL, python, and a machine learning engineering pipeline to deliver this.

Enhance this model with additional accuracy enhancements including a price. The model pipeline is set up for ongoing future iterations and improvements that we can continue to make over the long run for our business.

In the Linear TV space, both clearance and performance play critical roles in optimizing a clients campaign and allowing them to hit their budget and their efficiency goals. The performance prediction model has been rebuilt and developed on a machine learning pipeline, using SQL and python, and statistical methods. The model is now performing 2x better than historical versions and is set up for future interactions and improvements.

Streaming advertising accounts for a notably growing portion of Tatari's revenue and it will continue to grow as the industry evolves. The statistical model is developed end-to-end using a machine learning pipeline, SQL and python.

Tatari has an engine for both Linear and Streaming TV, in support of our major advertising channels that drive our revenue. These engines use the above mentioned predictive models and optimization methods. They also include data science features like built-in explainability. These products will continue to play a foundational role in the media buying optimization space, and require deep expertise in the underlying models, data, and engineering implementation.

Two Months Experience in Data Science or Data Analytics

Master's Degree in Data Science or a related field.

• Applying machine-learning techniques to real world problems.
• Using cloud computing tools and storage systems to develop data science solutions.
• Designing relational database models.
• Large-scale data-mining and machine-learning analysis
• Using Map Reduce framework.
• Manipulating large data sets.
• Programming in Python for retrieving, searching, and analyzing data from the Web.

F/T, 40 hours per week

3333 Michelson Dr, Suite 300, Irvine, CA 92612;
Hybrid Option: may work from home 2 days/week within commuting distance from company office in Irvine, Los Angeles, San Francisco or New York.

None.