SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 108,000 - 157,000 / annual
SeatGeek is seeking a Senior Data Analyst for the Risk Analytics team to serve as the analytical engine behind fraud prevention strategy. You will build statistical models, run experiments, and develop tools that reduce reliance on external black boxes and static rules, working closely with the Manager of Risk Analytics to translate strategy into measurable, executable outcomes.
Key responsibilities include building and maintaining Python-based analyses, models, and data pipelines supporting fraud decisioning and internal risk scoring. You will design and execute statistical experiments including A/B tests, holdout experiments, and vendor performance assessments. You'll develop and iterate on internal fraud risk models using SeatGeek transaction and vendor data, owning model calibration, validation, and ongoing performance monitoring.
A significant part of this role involves actively using AI tools—LLMs, code generation, and agentic workflows—to move faster and build smarter. You'll help define how AI gets embedded into the team's analytical processes and identify opportunities to automate manual work. You'll contribute to vendor performance analysis, assessing score calibration and measuring lift across segments to inform routing decisions and contract discussions.
You will build and maintain dashboards and reports in Looker and Hex, develop SQL models and data views, monitor fraud and operations metrics, investigate anomalies, and escalate findings with clear recommendations. Collaboration with Risk Ops agents, the manager, and cross-functional partners in Engineering, Payments, and Customer Experience is essential.
Required qualifications include 3+ years in fraud analytics, risk, fintech, or quantitatively demanding analytical roles. You must have strong Python skills (pandas, scikit-learn, statsmodels) and strong SQL for complex data pulls and warehouse work. Solid statistical grounding is essential—you should be able to design statistically valid experiments, perform significance testing, assess model calibration, and communicate findings to non-technical audiences. Hands-on experience building, training, and validating classification models independently is required, along with genuine enthusiasm for AI tools and comfort operating in ambiguity. Familiarity with fraud vendors (Forter, Riskified, Sardine) and BI tools like Looker is a plus.
