Pablo S.

→ Built Fabric data pipelines for sales and inventory reporting at one of Spain’s largest retailers, ingesting 7 source systems (~25 source objects) into OneLake with automated schema validation across bronze/silver/gold.

→ Built Fabric Warehouse data models for sales and inventory reporting, implementing incremental refresh strategies that reduced daily processing time by 35%.

→ Improved Power BI Direct Query performance by 47% through query optimization and view materialization in Fabric Warehouse, enabling sub-3s load times for executive dashboards.

→ Owned the credit-risk data domain at a major Spanish bank: Databricks lakehouse on Azure with PySpark + ADF medallion architecture, ∼2M daily records feeding regulatory reporting.

→ Built a Databricks (PySpark) data quality framework on Delta Lake (reconciliation, anomaly detection, schema and business-rule checks) gating merges into the credit-risk gold layer.

→ Automated operational workflows with Power Automate and SharePoint, replacing manual email-based approval processes for 4 business teams.

→ Migrated 12 legacy SQL Server databases to Azure SQL for a Spanish public sector entity, processing 50M+ records with automated validation and schema mapping.

→ Built ETL pipelines in Databricks to consolidate the 12 migrated databases, reducing daily refresh time from 3 hours to 45 minutes through partition pruning, optimized Delta writes, and broadcast joins.

→ Reduced semantic model size by 60% (1.6GB to under 650MB) for Fabric workspace compliance using Measure Killer and Tabular Editor, optimizing star schema relationships and eliminating unused measures.