About this role
Build root-cause ranking, incident classification and confidence-scoring models for Alarm Management, prepare labelled datasets and convert operator feedback into training signals. Deploy and monitor models in Amazon SageMaker with versioning, rollback and champion-challenger evaluation, combine ML predictions with deterministic rules, and ensure explainability and operational acceptance.
Skills for this role
Amazon SageMakerPythonscikit-learnXGBoostModel-serving APIsSupervised classificationAnomaly detectionClusteringHDBSCANConfidence scoringFeature engineeringLabelled dataset preparationModel evaluationPrecision and recallCalibrationFalse-positive rate analysisChampion-challenger evaluationMLOpsModel registryModel deploymentMonitoringDrift detectionRetrainingExplainabilityAlarm ManagementRoot-cause rankingIncident classification