Job DescriptionYou'll begin by diving deep into our factor usage: how factors are used, what they drive downstream, and how much they can move sim outputs. Early on, you'll design tests that catch unexpected changes to our simulations caused by factor moves. From there, we see this person consuming large volumes of market data from the exchanges and analyzing it to automate Contrarian Signals - adjusting our sims in response to market information. As we move into sports beyond the NBA, we'd expect this role to grow into building models for factor optimization. The work starts in analysis and moves into model building; it's lighter on infrastructure and software engineering, which is well covered by the team.
This position is remote from the USA or Canada.Duties:- Analyze our factor usage data - how factors are used, their downstream effects, and their impact on simulation outputs - and turn that analysis into actionable insight.
- Design and set up tests to detect unexpected changes to our sims resulting from factor moves.
- Ideate, develop, and improve machine learning and statistical models that drive Swish's core algorithms, growing into factor-optimization modeling as we expand to new sports.
- Develop contextualized feature sets that draw on sports-specific domain knowledge.
- Contribute across all stages of model development - from proof-of-concept and beta testing to partnering with data engineering and product teams to deploy new models.
- Constantly improve model performance using insights from rigorous offline and online experimentation.
- Assess model performance, identify weaknesses, and use those findings to direct development efforts.
- Document your work and present it clearly to technical and non-technical partners.
Requirements- Bachelor's in Data Science, Statistics, Computer Science, Applied Math, or a related technical field; Master's strongly preferred.
- 4+ years developing and delivering effective machine learning and/or statistical models to serve real business needs in sports analytics or sports betting.
- Strong foundation in Probability Theory, Machine Learning, Inferential Statistics, Bayesian Statistics, and Markov Chain Monte Carlo methods.
- Excellent analytical and problem-solving ability, and a demonstrated drive to learn quickly in unfamiliar territory.
- Experience with Python and relational SQL.
- Experience with source control (GitHub) and related CI/CD processes.
- Experience working in AWS environments.
- Ability to partner across teams on complex, ambiguous problems and communicate clearly with technical and non-technical audiences.
Base salary: Starting at $160,000 - DOE
Department Data Science Role NFL Team Locations San Francisco, CA - Remote Remote status Fully Remote
