Bogota, D.C., Capital District, Colombia
Led a team of 2 data scientists to design statistical models using AWS SageMaker for fraud detection on imbalanced datasets: - Achieved a 70% reduction in false positives Developed machine learning models (classification, regression, and time series) to optimize ATM operations: - Reduced operational costs by 40% - Improved customer experience by 60% - Increased cash availability in ATMs by over 50%
Designed and implemented ETL pipelines using: PySpark and Python Processed large-scale transactional data to feed analytics platforms and dashboards. Enabled high-performance data ingestion and transformation workflows supporting BI and ML projects.
Supported the BI team in early initiatives to analyze failed transactions and customer churn patterns. Collaborated on dashboard creation and performance indicators under guidance. Participated in the collection and structuring of transactional data for strategic use cases.