Machine Learning Engineer – Risk Scoring, Model Monitoring

This is a fully remote position, open to applicants in Latin America.

📋 Description

• Design and assess predictive models utilizing XGBoost, LightGBM, and CatBoost for risk scoring, classification, and associated decision-making support.

• Evaluate machine learning outcomes against deterministic rules and historical benchmarks.

• Identify shifts in input distributions, prediction behavior, and model accuracy, with notifications when performance dips below specified thresholds.

• Create feedback loops that leverage newly available outcome data, including retraining and validation processes.

• Recognize and mitigate target leakage, temporal leakage, proxy leakage, and unintended bias within model features.

• Establish reproducible model pipelines, encompassing monitoring, performance reporting, explainability, and controlled model deployment.

• Assess where machine learning contributes significant predictive value to a healthcare and pharmacy-finance platform.

• Communicate model performance, limitations, and business implications to both technical and non-technical stakeholders.

⛳️ Requirements

• Extensive production machine learning experience with Python.

• Practical knowledge of XGBoost, LightGBM, CatBoost, or similar gradient-boosted tree frameworks.

• Profound understanding of statistics, feature engineering, model evaluation, calibration, and temporal validation.

• Experience in detecting and managing data drift, concept drift, model degradation, and retraining triggers.

• Demonstrated ability to prevent data leakage and identify proxy variables that threaten model validity.

• Experience in developing production ML pipelines with automated monitoring, testing, and reproducibility.

• Ability to articulate model performance, limitations, and business implications to both technical and non-technical stakeholders.

• Background in healthcare, pharmacy, financial risk, credit scoring, or claims analytics.

• Familiarity with delayed outcome labels and historical transaction datasets.

• Experience in deploying interpretable models within regulated environments.

• Skills in integrating deterministic rules with probabilistic models.

🏝️ Benefits

• An opportunity to contribute to the establishment of a rigorous, measurable machine learning approach within a high-stakes business setting.