Data Scientist, NBA

About the Role & Team

We are looking for a Mid-Level Data Scientist to join our Data team. The Data Science team here at Penn is responsible for building models, forecasts, and simulations to help improve theScore Bet, ESPN Bet and iCasino products.  Our team values creativity, collaboration, ingenuity, and ownership.  We are looking for someone who is interested in joining a team responsible for building net new end-to-end sports data products to further our team’s mission of creating a world-class, data-driven product experience.
About the Work 

• Machine learning & statistical modeling: we want to be able to predict the likelihood or expected outcome of various events across different sports.
• AB Testing: we want data to inform the development of our products and features.
• Predictive & Prescriptive Analytics: we want to share in-depth knowledge on how our features are working and our forecasts are performing.
• Developing pipelines and developing models: we work closely with data and infrastructure engineers to deploy our models and develop required data pipelines to build data products.
• Develop best practices for our internal data processes that include model building, modeling techniques, improved latency, and readability.
• Design and build new predictive models and optimization routines that have an enterprise level impact. This varies from modeling expected sporting outcomes at an event level to utilizing various game state data to simulate full spectrums of expected outcomes.
• Collaborate with other members of the Data Science and Engineering teams on ways to approach problems, augment code, and share new techniques.
• Deploy modeling deliverables in conjunction with functional team leaders and stakeholders (in Product, Trading, etc.).
• Analyze results using solid statistical methods to iteratively improve data products.
• Communicate clearly, efficiently, and empathetically with technical and non-technical stakeholders.
• Write and maintain technical design and git/confluence documentation.
• Other duties as required.

About You

• Experience solving quantitative problems with MLB, NBA or NFL data.
• A firm grasp of utilizing sport-specific problem-solving techniques to model various elements of the game.
• The ability to dynamically project various aspects of player and team performance.
• Proficiency in developing simulations replicating the dynamics and intricacies of a sporting game.
• Proficient at writing code in Python and SQL to create meaningful insights.
• Experienced at creating algorithms, features, and applying machine learning data models to solve sports problems.
• Collaboratively iterative on data science products in an experimental fashion to rapidly adapt and improve products.
• Understanding of classification, regression and forecasting models, and A/B testing.
• Experience provisioning services using GitHub.
• Ability to take ownership of your work to achieve the necessary high-level objectives.

Nice to Have

• Familiarity with MLFlow.
• Familiarity with AWS/GCP.
• Experience setting up ML CI/CD pipelines, testing and validating code/data, managing databases, and deploying models.
• Experience with any of Docker, Kubernetes with Terraform, Cloudwatch.

What We Offer

• Competitive compensation package
• Education and conference reimbursements
• Parental leave top up
• Opportunities for career progression and mentoring others

#LI-REMOTE #LI-HYBRID