Antonio Rubí

Led the migration from SAP S/4HANA to BigQuery, co-designing a medallion-based Lakehouse architecture using Dataform and modular SQL logic. Developed a scalable transformation framework with dependency management and orchestrated workflows through Google Workflows and Cloud Scheduler. The implementation incorporated advanced optimization techniques such as partitioning, clustering, and table reorganization, achieving performance improvements between 50% and 70%.

Applied modeling techniques, including relational and dimensional design, creating one-big-table structures and star schemas optimized for performance and adapted to users usage patterns.

Collaborated with data owners to analyze distribution and implement array-based structures to optimize downstream consumption.

Additionally, integrated metadata governance using Dataplex and automated schema and table creation through Cloud Functions. Built internal Python tools to support testing, SQL linting, and metadata validation, improving reliability and standardization across the platform.

Designed a scalable, event-driven data architecture on AWS to process sensitive PII data from Parquet files (columnar data), using Apache Spark (PySpark) on AWS Glue for distributed batch processing. Aplplied optimization techniques such as broadcast joins, salting to mitigate data skew, repartitioning, caching…

Data was stored in partitioned Parquet format, improving performance and storage efficiency.

Processed outputs were made available internally via Athena and served externally and secured via API endpoints using Lambda and S3.

The orchestration layer combined AWS Step Functions and EventBridge for managing concurrent job execution, while lightweight Lambda functions were used to trigger Glue Spark jobs, handle API requests, and perform notifications-processing.

Metadata and schemas were managed via Glue Catalog while the infrastructure was defined and implemented in CloudFormation and AWS SAM, with CI/CD pipelines managed in GitHub Actions.

Integrated Snowflake into a centralized Lakehouse via Talend-based ETL pipelines, ingesting data from databases, SFTP, APIs, and message queues. Contributed to metadata governance and domain discoverability, designing one-table and star schemas based on usage. Gained hands-on experience with data lake architecture and warehouse tuning.

Designed scalable, cost-efficient data models in Snowflake, applying FinOps strategies that reduced monthly costs by 25%.

Applied dimensional design, creating one-big-table structures and star schemas. Optimized pipelines and queries through proper incremental designs, partitioning, clustering, and virtual warehouse tuning, using advanced SQL capabilities (CTEs, window functions, SCDs…).

Built the executive dashboard model using stored procedures and automated platform monitoring with tasks.

Leveraged key Snowflake features as time-travel, zero copy cloning, stages, and materialized views, …