Data Scientist — Model Quality

Find the signal in a million expert judgments: who grades well, which rubrics drift, and where quality quietly leaks.

What you'll do

• —Build statistical models of grader reliability and rubric health
• —Design sampling and audit strategies for expert-labeled data
• —Ship dashboards our operations team lives in
• —Chase anomalies until they confess

What we're looking for

• —Deep practical statistics (measurement, agreement, drift)
• —Strong SQL/Python and a bias toward shipping analysis
• —Experience with human-labeled data at scale
• —Healthy suspicion of clean-looking metrics