Mariano Cruz

Owned end-to-end data management for Marketing & Sales SaaS, including building and maintaining datasets, data pipelines, and canonical, audit-ready data models with documented business rules. Orchestrated and monitored scheduled imports and data loads with Apache Airflow, applying pre/post-load QA and error handling across multiple sources. Standardized batch pipelines using dbt-style transformation patterns and incremental load strategies, reducing onboarding time for new data sources by 50%. Automated anomaly detection and root-cause analysis in Python (pandas) across 10+ datasets and improved data incident resolution time by 35%. Built self-serve dashboards and modular datasets in AWS QuickSight, reducing ad-hoc requests by 8+ hours/week, and supported PII governance for customer/contact data (email permission updates, entity identification and tagging). Maintained team Playbooks, SOPs, naming conventions, and data dictionaries to make repeatable processes and reduce human error.

Optimized 20+ SQL/PLSQL queries and execution plans, reducing critical KPI report generation time by 45% across 3 business units and supporting senior leadership decision-making. Achieved 99.8% batch pipeline uptime by redesigning scheduled loads and orchestrating/monitoring jobs with Control-M, reducing business-impacting incidents by 60%. Resolved major ETL/ELT bottlenecks on nationwide datasets (millions of records) through profiling, restructuring, and partitioning improvements, saving 12+ analyst hours per week. Built Power BI dashboards from scratch for 5 non-technical teams, enabling self-serve reporting and reducing analyst backlog by 40%. Translated transactional data into findings and recommendations for both technical and non-technical stakeholders.