About this role
Build ML systems that score, validate, and improve complex work products with noisy, delayed, or disputed labels; design evaluation frameworks and feedback loops to measurably improve model and system behavior. Own production ML behavior end-to-end (precision/recall, regression detection, drift, latency, cost, explainability) and partner with backend engineers to integrate inference into long-running workflows.
Skills for this role
Machine LearningEvaluation designError analysisDebugging production ML failuresPromptingFine-tuningRetrievalActive learningHeuristicsHuman-in-the-loop MLLLM applicationsModel-assisted workflowsPythonTemporalPostgresAWSLiteLLMRegression detectionDrift monitoringExplainabilityPrecision/recall tradeoffsModel inference integrationHuman review