Senior Data Scientist (Remote)

Key Skills

Java Programming, Python Programming, SQL, AWS, NumPy, Pandas, Graylog, PySpark, Kubeflow, MLflow, Seldon

Job Description

Paramount Global (NASDAQ: PARA, PARAA) is a leading global media and entertainment company that creates premium content and experiences for audiences worldwide. Driven by iconic studios, networks and streaming services, Paramount's portfolio of consumer brands includes CBS, Showtime Networks, Paramount Pictures, Nickelodeon, MTV, Comedy Central, BET, Paramount+, Pluto TV and Simon & Schuster, among others. Paramount delivers the largest share of the U.S. television audience and boasts one of the industry's most important and extensive libraries of TV and film titles. In addition to offering innovative streaming services and digital video products, the company provides powerful capabilities in production, distribution and advertising solutions.

About the Content Optimization Team

We are looking for a Data Scientist to join the newly crafted Content Optimization team. Our mission is to use modern data science to inform decision-making relating to content windowing and investment across ViacomCBS businesses. In this role, you will join a group of top-notch data scientists and data engineers who build predictive models that support highly critical multi-million dollar content-related decisions. This will be achieved in collaboration with our other data science and research groups across ViacomCBS.

The role reports to the SVP of Content Data Science. This represents a uniquely exciting opportunity for a data scientist to deliver impact at a leading premium global content company, in an exciting and rapidly-evolving industry.

We are seeking individuals with deep technical skills who seek to be the smartest in the room and are not afraid of being surrounded by smarter ones, individuals who are results-oriented enough to work on multi-year projects that have the potential to build significant value across all parts of our business.

About you. You must be:

• Feel very comfortable asking a lot of questions
• Comfortable with Kanban and Scrum as well as no structure at all because you understand that one model does not fit all 
• Apt at being at time a leader and at other time a follower, possibly with the same people regardless of ranks and seniority, depending on how you contribute to a project
• Enjoy written communication because it gives you a medium to express highly complex and nuanced ideas

Overview & responsibilities:

• Preference is to be based in Atlanta, GA but may work remotely within the US
• The lead data scientist uses advanced statistical and scientific methods to deliver insightful analysis that helps to inform critical business problems relating to content valuation
• This role reports to the Senior Vice President for Content Data Science

This is a role responsible and accountable for:

• Building models (e.g. acquisition, retention, viewing behavior on streaming platforms)
• Servicing multiple ad hoc requests in a timely manner
• Ensuring regularity and standardization in data, metrics, and analysis

Basic qualifications:

• 6+ years of commercial experience in a data scientist role, capable of writing production grade code
• Knowledge of current ML, AI techniques, both in terms of implementation and theory
• MLOps experience (MLFlow, Kubeflow, Seldon)
• Excellent Python (including Pandas and NumPy), familiarity with TensorFlow
• Excellent knowledge of OOP, functional programming, and familiarity with large code
• Experience with Sagemaker
• Knowledge of distributed frameworks (PySpark, Ray, Modin for instance)
• Familiarity with SQL dialects
• Familiarity with AWS environment and deployment
• Experience with large volume of data (100 tb) as well as with tiny ones (n=50)
• Experience with microservices architecture
• Master degree (PhD preferred) in Data Science, Computer Science, Statistics, Math, Physics, Astronomy, Economics or equivalent

Desired additional qualifications:

• Scientific or technical publications
• Experience with Random Forest and Boosting
• Comfortable questioning and challenging assumptions, processes to improve outcomes
• Versatile (capable of using anything from rank statistics to shape theory to topological data analysis to the most obscure time series model)

#LI-AW3