Christian Rodriguez

Served as a primary focal point for pipeline development and data distribution, managing 15+ automated pipelines for 50+ internal users. Led dashboard development (Power BI/Tableau), cutting manual reporting time by ~35% and increasing positive donor feedback scores by 20% through semi-automated donor reporting.

Built risk-scoring models in Python and R to identify high-risk operations, reducing undetected anomalies by ~25%. Designed ETL processes across 8+ corporate data sources, cutting reconciliation time by 30% and enabling near-real-time reporting for 10+ regional offices.

Conducted data assessments delivering 10+ analytical reports that informed operational improvements. Integrated 12+ heterogeneous data sources, building automated Python/R pipelines that reduced data preparation time by 40% and decreased data inconsistencies by ~30% via validation rules.

Interpreted complex financial and risk data to produce underwriting recommendations covering 500+ monthly insurance policies. Built a Python/PostgreSQL ETL system and developed fraud detection and credit risk predictive models, improving fraud identification by ~15%.

Led a regulatory inventory initiative using NLP and machine learning to classify 5,000+ legal documents and support anomaly detection for high-risk urban areas adopted by 3 municipal governments. Built R Shiny dashboards for 10+ government analysts to enable evidence-based policymaking.

M.Sc. in Artificial Intelligence with a thesis on an emotion-driven music recommendation system using ensembled CNNs for facial emotion classification and a Multi-Armed Bandit reinforcement learning approach. Final ensemble accuracy reported as 66.69%.

B.Sc. in Statistics from the Faculty of Basic Sciences.